National Register Eligibility Testing of 41MM340 and 41MM341, along Little River, Milam County, Texas

by Richard B. Mahoney, Steve A. Tomka

with Contributions by J. Philip Dering, Richard W. Fullington, Robert G. Howells, John G. Jones, Raymond P. Mauldin, Barbara A. Meissner, Lee C. Nordt, Timothy K. Perttula, and Barbara Winsborough

Center for Archaeological Research Environmental Affairs Division The University of Texas at San Antonio Archeological Studies Program, Report No. 30 Archaeological Survey Report, No. 303

March 2001 ISBN: 1-930788-11-8 National Register Eligibility Testing of 41MM340 and 41MM341, along Little River, Milam County, Texas

by Richard B. Mahoney and Steve A. Tomka

with Contributions by J. Philip Dering, Richard W. Fullington, Robert G. Howells, John G. Jones, Raymond P. Mauldin, Barbara A. Meissner, Lee C. Nordt, Timothy K. Perttula, and Barbara Winsborough

Texas Antiquities Committee Permit No. 2319

Robert J. Hard and Raymond P. Mauldin Principal Investigators

Prepared for: Prepared by: Texas Department of Transportation Center for Archaeological Research Environmental Affairs Division The University of Texas at San Antonio Archeological Studies Program, Report No. 30 Archaeological Survey Report, No. 303

March 2001 National Register Eligibility Testing of 41MM340 and 41MM341, along Little River, Milam County, Texas

Copyright ©2001 Texas Department of Transportation (TxDOT) and Center for Archaeological Research, The University of Texas at San Antonio (CAR–UTSA)

All rights reserved

TxDOT and CAR–UTSA jointly own all rights, title, and interest in and to all data and other information developed for this project. Brief passages from this publication may be reproduced without permission provided that credit is given to TxDOT and CAR–UTSA. Permission to reprint an entire chapter, section, figures or tables must be obtained in advance from the Supervisor of the Archeological Studies Program, Environmental Affairs Division, Texas Department of Transportation, 125 East 11th Street, Austin, 78701. Copies of this publication have been deposited with the Texas State Library in compliance with the State Depository requirements.

Printed by Thomas Reprographics Inc., San Antonio, Texas March 2001

jointly published by

Texas Department of Transportation Environmental Affairs Division Archeological Studies Program Nancy A. Kenmotsu, Ph.D., Supervisor

Archeological Studies Program, Report No. 30 A. McGraw, Series Editor

and

Center for Archaeological Research The University of Texas at San Antonio Archaeological Survey Report, No. 303

Printed on acid-free, 60 lb. paper ISBN: 1-930788-11-8 Abstract From January through March 2000, the Center for Archaeological Research, The University of Texas at San Antonio conducted National Register of Historic Places eligibility testing for archaeological sites 41MM340 and 41MM341, under contract with Texas Department of Transportation. The investigations were conducted under Texas Antiquities Permit Number 2319. The Phase II testing fieldwork consisted of excavation of 20 test units across both sites to investigate significant cultural deposits encountered during the previous survey phase. Four distinct stratified cultural zones at 41MM340 and at least two cultural zones at 41MM341 were identified during the testing efforts. In concert with the archaeological field investigations, the following special analyses and studies were performed to aid the determination of site integrity and eligibility: geoarchaeology, radiocar- bon, lithic, aboriginal ceramic, vertebrate faunal, freshwater mussel shell, macrobotanical, pollen, phytolith, land snail shell, amino acid racemization, and diatom. The synthesis of these analyses has provided adequate data to determine temporal integrity and recommendation of National Register eligibility for both sites. Further cultural resources investigations in the form of Phase III data recovery excavations are thus recommended for both sites prior to development.

i Management Summary National Register of Historic Places eligibility testing for archaeological sites 41MM340 and 41MM341 in Milam County, Texas was conducted by the Center for Archaeological Research, The University of Texas at San Antonio from January through March 2000. The testing was performed for the proposed Texas Department of Transportation development of State Highway 36 near the city of Cameron. Construction plans for the pro- posed development include erection of twin bridges, replacing the extant 1937 single bridge, spanning the 1.2 km floodplain of Little River. Adverse impact resultant from construction within the current and proposed additional 21 m rights-of-way would destroy all or portions of these two prehistoric archaeological sites.

The current cultural resources testing phase consisted of the excavation of 20, 1-m2 test units across the two sites. At least six distinct occupation surfaces have been identified, ranging from the Late Archaic I Period to the Late Prehistoric Period (ca. 1050 B.C. to A.D. 1450). Numerous stone tools and aboriginal ceramics were recov- ered and analyzed. By means of radiocarbon assays and analyses of temporally diagnostic artifacts, both sites have exhibited relatively high temporal integrity. As such, it is recommended that further cultural resources investigations in the form of Phase III data recovery excavations be performed prior to proposed development.

ii Contents Abstract...... i Management Summary ...... ii Figures ...... iv Tables...... v Acknowledgments ...... vi Chapter 1: Introduction...... 1 Project History ...... 1 Chapter 2: Cultural Setting ...... 3 Aboriginal Cultural Setting ...... 3 Historic Cultural Setting ...... 4 Chapter 3: Environmental Setting ...... 7 Paleoenvironmental Conditions ...... 8 Chapter 4: Archaeological Background...... 11 Previously Recorded Sites...... 11 Chapter 5: Geoarchaeology ...... 13 Geology and Geomorphology ...... 13 Chapter 6: Methodology ...... 21 Mechanical Excavations...... 21 Manual Excavations ...... 22 Chapter 7: Results ...... 27 41MM340 ...... 27 41MM341 ...... 49 Chapter 8: Recommendations...... 59 Site Eligibility ...... 59 Chapter 9: Data Recovery Plan ...... 63 Data Types and Research Issues...... 63 References Cited...... 71 Appendix A Soil - Stratigraphic Descriptions ...... 76 Appendix B Radiocarbon Results...... 86 Appendix C Vertebrate Faunal Remains ...... 90 Appendix D Freshwater Mussel Shell ...... 94 Appendix E Macrobotanical Analysis of Flotation Samples ...... 100 Appendix F Pollen and Phytolith Results ...... 104 Appendix G Gastropod Analysis ...... 106 Appendix H Diatom Paleoenvironmental Analysis of Sediments ...... 114 Appendix I Amino Acid Racemization Values for Snail and Mussel Shell ...... 120 Appendix J Prehistoric Ceramic Sherds from 41MM341 ...... 124

iii Figures Figure 1-1. General project area showing Little River and associated floodplain...... 2 Figure 3-1. Color infrared aerial photo of general project area...... 7 Figure 3-2. Project area in relationship to Natural Regions of Texas...... 8 Figure 5-1. Little River Basin, Milam County, Texas...... 14 Figure 5-2. Alluvial stratigraphic cross-section on the Project Area...... 15 Figure 5-3. Detailed soil-stratigraphic profiles from backhoe localities...... 17 Figure 5-4. Schematic alluvial stratigraphic cross-section of the Project Area...... 19 Figure 6-1. EDM-derived topographic map of 41MM340...... 23 Figure 6-2. EDM-derived topographic map of 41MM341...... 24 Figure 7-1. South wall of excavation Unit C, Block 2, at 41MM340 with zones identified...... 28 Figure 7-2. Profile of south wall of excavation units C and D, Block 2, 41MM340...... 29 Figure 7-3. Profile of south wall of excavation units C and A, Block 3, 41MM340...... 29 Figure 7-4. Excavation units A, B, C, and D, Block 2, 41MM340...... 30 Figure 7-5. Distribution of chipped stone for all excavation units in Block 1 and Unit B, 41MM340...... 30 Figure 7-6. Distribution of chipped stone for all excavation units in Block 2 and Unit A, 41MM340...... 30 Figure 7-7. Standardized fire-cracked rock and mussel shell weights by zone, Block 2, 41MM340...... 31 Figure 7-8. Standardized volume of chipped stone recovered by zone in Block 3, 41MM340...... 31 Figure 7-9. The vertical provenience of carbon samples for Block 2, Unit C, 41MM340...... 33 Figure 7-10. Probability distributions of calibrated radiocarbon dates for Block 2, 41MM340...... 34 Figure 7-11. Probability distributions of calibrated radiocarbon dates for Block 3, 41MM340...... 34 Figure 7-12. A/I values for Fort Hood proveniences and 41MM340...... 36 Figure 7-13. Total snail weight (gr.) by zone for Block 2 and Block 3 at 41MM340...... 38 Figure 7-14. A/I ratios (less than 0.2) for mussel shells by zone at 41MM340...... 38 Figure 7-15. Dart points and Erath biface recovered from 41MM340...... 40 Figure 7-16. Feature 2, Block 3, 41MM340...... 42 Figure 7-17. Feature 4, Block 2, Unit C, 41MM340...... 42 Figure 7-18. South wall of BHT 24 with occupation zones and dates of radiocarbon assays, 41MM341...... 50 Figure 7-19. East wall of Unit D, Block 2 with occupation zone and date of radiocarbon assay, 41MM341. . 51 Figure 7-20. Chipped stone tools recovered from 41MM341...... 52 Figure 7-21. Distribution of chipped stone items by zone in Block 1 at 41MM341...... 55 Figure 7-22. Distribution of chipped stone items by zone in Block 2 at 41MM341...... 55

iv Tables Table 4-1. Previously recorded Prehistoric Archaeological Sites along Little River in Milam County ...... 11 Table 7-1. Beta Analytic, Inc., results of radiocarbon dating for samples recovered from 41MM340 ...... 32 Table 7-2. Radiocarbon interval estimates and deposition rates, 41MM340 ...... 35 Table 7-3. List of potential chronologically diagnostic artifacts recovered from 41MM340 ...... 39 Table 7-4. Features encountered during manual excavation at 41MM340 ...... 41 Table 7-5. Vertical Distribution of Artifacts and Ecofacts Recovered from Block 1, 41MM340 ...... 43 Table 7-6. Vertical Distribution of Artifacts and Ecofacts Recovered from Block 2, 41MM340 ...... 44 Table 7-7. Vertical Distribution of Artifacts and Ecofacts Recovered from Block 3, 41MM340 ...... 44 Table 7-8. Standardized Artifact and Ecofact Distribution, Block 2, 41MM340 ...... 45 Table 7-9. Standardized Artifact and Ecofact Distribution, Block 3, 41MM340 ...... 45 Table 7-10. Comparison of debitage cortex categories, Block 2, 41MM340...... 46 Table 7-11. Comparison of debitage cortex categories, Block 3, 41MM340 ...... 46 Table 7-12. Comparison of the occurrence of three flake types by zone, Block 3, 41MM340...... 47 Table 7-13. Beta Analytic, Inc., results of radiocarbon dating for samples recovered from 41MM341 ...... 50 Table 7-14. Distribution of chipped stone tools and ceramics by block, unit, zone, and level, at 41MM341 .. 51 Table 7-15. Features encountered during manual excavation at 41MM341 ...... 53 Table 7-16. Vertical Distribution of Artifacts and Ecofacts in Block 1, 41MM341 ...... 54 Table 7-17. Vertical Distribution of Artifacts and Ecofacts in Block 2, 41MM341 ...... 54 Table 7-18. Percentage of cortex categories in Block 1, 41MM341 ...... 56 Table 7-19. Percentage of cortex categories in Block 2, 41MM341 ...... 56 Table C-1. Taxa identified from 41MM340 and 41MM341 ...... 91 Table D-1. Identification of freshwater mussel shell remains recovered from 41MM340 and 41MM341 ..... 95 Table D- 2. Freshwater mussel assemblages documented at locations in the Central Brazos River Basin ...... 96 Table E-1. Plant taxa identified in the samples ...... 101 Table E-2. Flotation samples, provenience, liters floated, and identifications...... 101 Table G-1. Identification of gastropod shell remains recovered from 41MM340 ...... 107 Table G-2. Distribution of selected gastropod shell remains recovered from 41MM340...... 108 Table G-3. Identification of gastropod shell remains recovered from 41MM341 ...... 110 Table G-4. Distribution of selected gastropod shell remains recovered from 41MM341...... 111 Table H-1. Samples recovered from 41MM340 and 41MM341 examined for diatom content ...... 115 Table H-2. Identification of diatom samples recovered from Little River and 41MM340 and 41MM341 .... 116 Table I-1. Snail Shell Ratios and ASP/GLU Ratios ...... 120 Table I-2. Mussel Shell A/I Ratios and ASP/GLU Ratios ...... 121

v Acknowledgments Numerous individuals and organizations contributed to the success of this project, foremost among these were the field crews. During the field effort, Steve A. Tomka and Richard B. Mahoney of CAR served as Project Archaeologists. Crew members, who worked well under a variety of less than hospitable conditions included Donna Edmondson, Antonia Figueroa, Owen Ford, Brian Langner, Preston McWhorter, Barbara A. Meissner, and Brandon S. Young. Consultants on the project included Dr. J. Philip Dering, Richard W. Fullington, Robert G. Howells, John G. Jones, Lee C. Nordt, Timothy K. Perttula, and Barbara Winsborough. Steve Tomka performed the lithic analysis and Barbara Meissner analyzed the faunal material. In addition, TxDOT archaeologists Steve Ahr and James Abbott provided invaluable assistance. Bruce Moses and Richard Young of CAR drafted the excellent illustrations in the report, while Maryanne King and Johanna Hunziker, CAR editors, produced this final report. Myles Miller of THC provided critical advice on the radiocarbon dates, while Dr. G. Lain Ellis of TxDOT contributed valuable information on interpreting the snail and mussel racemization data. Special thanks to Shannon Roberts of D&M Septic, Groesbeck, Texas; Sam Sosa of Varsity Motel, Cameron, Texas; and Jim Carter of Carter Backhoe, Cameron Texas.

vi Chapter 1: Introduction

The Center for Archaeological Research (CAR), The corridor. The resultant twin bridges will each measure University of Texas at San Antonio was contracted by 14.9 m (49 ft.) in width and will be separated by a Texas Department of Transportation (TxDOT), Austin, distance of 22.8 m (75 ft.). Texas, to conduct test excavations for two previously recorded prehistoric archaeological sites. The purpose of the current investigations was to assess the National 1998-1999 Cultural Resources Investigations Register and State Archaeological Landmark eligibility of sites 41MM340 and 41MM341, located within the Archaeological sites 41MM340 and 41MM341 were present and proposed highway rights-of-way, prior to discovered and recorded in 1998 by TxDOT the proposed expansion of State Highway 36 at archaeologists, Steve Ahr and Jim Abbott, during a Cameron, Milam County, Texas. Concurrent with the survey of the approximately 2.52 ha (6.23 ac.) of the archaeological investigations, geomorphological studies Little River floodplain. The survey phase consisted of were conducted by Dr. Lee C. Nordt of Baylor excavation of 28 backhoe trenches placed at roughly University. The investigations were conducted under 50 m (164 ft.) intervals across the approximately Texas Antiquities Permit Number 2319, with Dr. Robert 1.2 km (0.75 mi.)-wide floodplain. Radiocarbon dating J. Hard, former CAR Director, and Dr. Raymond of samples extracted from various features during the P. Mauldin, CAR Assistant Director, serving as survey phase suggest a date of ~1050 BP (Late co-principal investigators. Sites 41MM340 and Prehistoric) for the single component encountered at 41MM341 are located approximately 400 m (1312 ft.) 41MM341 and ~2800 BP (Late Archaic) for one and 600 m (1969 ft.) inland, respectively, along the left component of 41MM340. descending bankline of Little River in the central portion of Milam County, Texas (Figure 1-1). 2000 Cultural Resources Investigations

Project History National Register eligibility assessment testing for sites The portion of the proposed construction affecting the 41MM340 and 41MM341 was conducted by CAR Project Area is part of a more extensive TxDOT archaeologists from January through March 2000. effort entitled, The Proposed State Highway 36 Testing efforts consisted of mechanical excavation of Bridge Replacement over the Little River and Road 14 backhoe trenches and manual excavation of 20 2 Widening between Travis Street in Cameron to 1-m test units within the previously delimited site US 77, Milam County (TxDOT CSJ: 0185-04-033 & boundaries. The results of these investigations are 034). The bridge replacement aspect of the project contained herein. will extend the western right-of-way 21 m (70 ft.) and maintain the current eastern right-of-way, with the western extension the focus of the cultural resources Report Layout investigations. The first phase of development will involve construction of a new bridge west of the The report is divided into nine chapters with 10 existing bridge, ultimately providing a southbound appendices. Following this introductory chapter, the corridor. The second phase of development involves Cultural Setting chapter provides a brief overview of the demolition and subsequent reconstruction of the the cultural prehistory and history of the region. The current bridge, providing the accompanying northbound third chapter, Environmental Setting, discusses the

1 Figure 1-1. General project area showing Little River and associated floodplain. Cameron, TX USGS topographic quadrangle (7.5' series) 1961 (photograph revised 1982).

general physical environment encountered within the The ten appendices provide supporting data for the Project Area. Chapter Four, Archaeological analysis. Appendix A provides detailed soil and Background, presents previous archaeological stratigraphic descriptions. Appendix B contains investigations within the region and an overview of radiocarbon data sheets. Appendix C presents data on previously recorded sites. Geoarchaeology, located in faunal material. Appendix D presents freshwater Chapter Five, discusses the geomorphology and mussel data. A macrobotanical summary is presented geoarchaeological interpretations of the Project Area. in Appendix E, while pollen and phytolith results are The sixth chapter describes, in detail, the field discussed in Appendix F. Appendix G discusses land methodology employed during the investigations, site snails recovered from the two sites, and diatoms are descriptions, feature descriptions, and artifact discussed in Appendix H. Appendix I presents amino descriptions. The seventh chapter discusses the results acid racemization results for snails and mussels, and from the field and laboratory investigations at both sites. Appendix J discusses the ceramics which were The eighth chapter presents recommendations for recovered from 41MM341. further work, and Chapter 9 provides a detailed data recovery plan to be implemented in subsequent mitigative efforts. 2 Chapter 2: Cultural Setting

The cultural setting of the Little River floodplain and The commencement of the Late Archaic I phase the surrounding region is here addressed relative to relative to the Project Area is characterized by a the chronology exhibited by the radiocarbon samples generally xeric environment probably correlative with and associated diagnostic artifacts of the cultural the Dry Edwards Interval to the west and southwest. occupation zones. Specifically, as the oldest component This drying trend is evidenced by an overall increase encountered falls within the Late Archaic I Period, in reliance on grassland species as opposed to aquatic that is where our description of the Aboriginal Cultural resources among numerous sites in the North Fork Setting will begin. A more detailed account of the Reservoir vicinity (Peters et al. 1982:16-29). Johnson various prehistoric periods and phases here discussed, and Goode propose that, due to the xeric conditions as well as the chronology of the Central Texas region experienced by the peoples of the Late Archaic I period, in toto, can be found in Collins (1995), Johnson (1995), burned rock middens proliferate for the processing of Johnson and Goode (1994) and Prewitt (1981). semi-succulents (1994:34-35). Additionally, the period is further defined by the projectile-point styles of the Bulverde, Pedernales, Marshall, Montell, and Aboriginal Cultural Setting Castroville (Johnson and Goode 1994:Figure 2). Archaeologically, this interval is subdivided into two Johnson and Goode suggest Eastern (United States) generally accepted time periods, the Late Archaic and religious influences, manifest in the form of various the Late Prehistoric. The further delineation of these burial practices, as one of the primary indicators of periods into sub-periods or phases remains a perennial the Late Archaic II phase (1994:37). The continuum topic of debate within the archaeological community. of the trend toward a mesic environment can also be Accordingly, this brief discussion will utilize a synthesis attributable to this period change. While a definitive of Johnson and Goode’s breakdown of the Late date cannot be placed upon the abandonment of burned Archaic (1994) and Prewitt’s subdivision of the Late rock middens, Johnson and Goode note that these Prehistoric (1981). feature types are generally associated with the Late Archaic I phase, and the absence thereof denotes The Late Archaic period represents the final three the beginning of the Late Archaic II phase (1994). millennia of the Archaic Era, from approximately Typical projectile-point styles of this phase include, in 4200 BP to 1200 BP (Johnson and Goode 1994:29). progressive order, Marcos, Ensor, Frio, Darl, and Adaptation to a relatively dry climate with low Figueroa (Johnson and Goode 1994:Figure 2). precipitation and high temperatures appears to mark the beginning of the period, with bison reappearing in Transition from the Late Archaic to the Late Prehistoric the faunal assemblage following an over one-thousand- is arguably accepted to occur with the advancement year hiatus (Dillehay 1974). Despite these xeric in technology from hunting techniques utilizing the atlatl conditions, human population seems to have increased and dart to utilization of the bow and arrow. Additionally, within the region (Prewitt 1985 in Johnson and Good the occurrence of ceramic vessels generally denotes 1994:36). Investigations conducted at the Chesser Site this change of periods. The Late Prehistoric period is (41LE59), a multi-component site spanning the Paleo- here defined as the time frame of approximately Indian through Late Prehistoric, suggest a substantial 1200 BP until European contact, roughly 250 BP growth in population with the onset of the Late Archaic within this region. period, based upon increases of stone tools and lithic debitage (Rogers and Kotter 1995:134).

3 Prewitt identifies the initial succeeding Late Prehistoric de Horcasitas. The mission was established, in part, to phase as the Austin Phase, occurring from the attract the Tonkawa (or Titskanwatits); however, by termination of the Late Archaic II until approximately the 1740s the Tonkawa had migrated only as far south 650 BP (1981:Figure 3). Aside from the aforementioned as the Red River Valley and were not drawn to the changes in technology, Prewitt ascribes only a slight mission (Prikryl 2001). As local interest in the mission increase in the dependence upon hunting as a means increased, two more missions, San Ildefonso and of subsistence and a marked increase in the occurrence Nuestra Señora de la Candelaria, were established in of “true cemeteries” as an indicator of period change 1749. The missions attracted a diverse population, and, (1981:74). The artifact assemblage recovered from the as such, each mission catered to a different linguistic Loeve-Fox Site (41WM230) suggests gathering takes group: San Francisco Xavier de Horcasitas contained precedence over hunting during the Austin Phase primarily Mayeye; San Ildefonso contained Atakapan, (Prewitt 1982:179). Representative artifacts include including Akokisa, Bidai, and, Deadose; Nuestra Scallorn points and Friday bifaces. Señora de la Candelaria contained, interestingly enough, Karankawan, primarily Coco (Chipman 1992). The relatively short-lived Toyah phase, as defined by Prewitt (1981), is characterized by the “dramatic” shift Although occasionally raided by the neighboring Lipan in subsistence from hunter-gatherer to that of an Apache, the missions thrived and were granted a economy based primarily upon hunting. Based upon presidio in 1751 by Viceroy Conde de Revilla Gigedo. data from Dillehay (1974), bison once again reappear The establishment of the new presidio, San Francisco in the faunal assemblage of archaeological sites within Xavier de Gigedo, brought with it the new regime of a Central Texas. An intermediate shift to a generally dry, garrison under the command of Captain Felipe de mesic environment is attributed to this influx of ungulate Rábago y Terán. Rábago’s disdain for the missionaries dependence (Johnson 1995). The material culture of and neophytes was reflected in various incidents, one this time period, in the form of various stone tools of which resulted in the excommunication of the entire utilized for bison procurement and processing such as garrison. Due primarily to the chaos and instability of Edwards, Perdiz, and Scallorn arrow points, along with Rábago’s tyranny, the neophytes ultimately abandoned various scrapers and other stone tools, appears the complex, and in 1755 the presidio was transferred to reflect a subsistence based upon the procurement to the San Marcos River. of bison. In 1825, the Mexican government issued a colonization grant to the Texas Association of Nashville, Tennessee. Historic Cultural Setting Failure to promote the settlement of the colony lead to a string of subsequent grantees of the Mexican Spanish exploration through present Milam County was government, including Sterling Robertson and first chronicled during the expedition of Francisco Isidro Stephen F. Austin. In his second bid for control of the Felix de Espinosa and Domingo Ramón in 1716. Upon grant, Robertson finally succeeded in colonization, and encountering Little River and its primary tributary, the settlement of the area began. Initially named the San Gabriel River, Ramón named the streams San Municipality of Viesca, the Republic of Texas renamed Andres and San Xavier, respectively. These names the district Milam County in honor of Benjamin would persist until anglicized versions ultimately Rush Milam in 1836. replaced them during the nineteenth century. The population of Milam County consistently rose over Representatives from the Deadose and Mayeye tribes the following decades with an agricultural based approached the Spanish government in the 1740s to economy. However, similar to a majority of agricultural- request placement of a mission along the San Gabriel based communities in the South, the Civil War and River near the confluence of Brushy Creek. Mariano subsequent Reconstruction placed its toll upon Milam Francisco de los Dolores y Viana was dispatched and County. It would not be until the mid-1870s, when the by 1746 had established Mission San Francisco Xavier International and Great Northern Railroad was

4 constructed, providing a major artery through the county, connecting Austin and the Brazos River, a more efficient means by which to export locally produced crops was realized. Similarly, in the early 1880s and 1890s, the county was further serviced by the Gulf, Colorado, and Santa Fe Railroad and the San Antonio and Aransas Pass Railroad, respectively.

Within the past century, Milam County has undergone a varied swing in population and industrial technological diversions. In the 1890s, lignite mining became a viable natural resource of the county, providing an economic alternative to the standard of agriculture in the area. Additionally, a short boom of oil production occurred in the 1920s. Both of these industries were short-lived, however, and the population soon began a sharp decline as workers left for more productive areas. The Aluminum Company of America (Alcoa) re-engineered the process for application of lignite and, in the 1950s, revived lignite mining in the county. Currently, Milam County still relies heavily upon agriculture as an economic base, with industries providing a substantial boost to the once impoverished county.

5

Chapter 3: Environmental Setting

The Project Area is located in the central portion of and on the banks of the stream afforded diverse, Milam County, Texas, approximately 1.2 km (0.75 mi.) seasonal resources. Indeed, Spanish exploration of the southeast of the city of Cameron. The two sites are area in the mid-1700s indicated an abundance of bison, located within a floodplain approximately 500 m deer, turkey, fish, and wild fruits (Gilmore 1969). (1641 ft.) inland along the left descending bankline of Little River. As depicted on current USGS According to recent, unpublished Soil Conservation topographic maps and aerial photographs Service (SCS) soil survey maps of the Project Area, (Figure 3-1), the expansive floodplain is dissected by 41MM340 is located within the Tinn series of very numerous sloughs, representing relict meander scars deep, moderately well drained, very slowly permeable of the associated stream. soils that formed in calcareous clayey alluvium. The site sits on an apparent point bar formed by a former The sites are located at the border of the Blackland meander of Little River. This meander is still visible Prairie and Post Oak Savannah vegetation regions in the vicinity of the site and is part of a discontinuous (Figure 3-2). Little River divides these two regions at line of sloughs, backwater ponds, and Green Lake. this location. Prairie grasslands atop clayey soils A slight escarpment (~50 cm high) located along the predominate along the left descending bankline. The probable former terrace is only 100 m (328 ft.) south location of the sites at the juncture of these two regions of the site.

Figure 3-1. Color infrared aerial photo of general project area.

7 Figure 3-2. Project area in relationship to Natural Regions of Texas.

SCS maps within this soil survey depict 41MM341 ragweed (Ambrosia trifida) thickets, colloquially within the Frio series of very deep, well-drained, referred to as “bloodweed.” A riparian zone exists moderately slowly permeable soils that formed in along the bankline of Little River consisting of various loamy and clayey calcareous alluvium. Classified as hardwoods, primarily pecan (Carya illinioensis). a Mollisol, the silty clays exhibit typical vertic Cattle are perennially maintained along the western properties of frequent shrink-swell dynamics. Indeed, section of the proposed right-of-way and seasonally during the testing phase, vertic cracks approximately introduced into the remainder of the right-of-way. 10 cm in width were noted at ground surface on both sites, and evidence of the fissures was encountered at terminal excavation depths of 1.6 m (5.25 ft.). It should Paleoenvironmental Conditions be noted that vertic properties are not characteristic of Frio soils, and the correct assignment of soil The reconstruction of the paleoenvironmental classification would more likely be Denton or Slodell conditions characterizing the vicinity of the two Little soils, which are commonly associated with Frio soils. River sites is based on pollen records from a number of localities in close proximity to the sites. These The area within the present and proposed rights-of- include: Gause Bog (Milam County), Boriack Bog and way is currently used as pasture land. The predominant Patschke Bog (Lee County), Franklin Bog (Robertson vegetation is ground-cover grasses and dense giant County), and Weakly Bog (Leon County). Since the

8 three bogs are located along the western edge of the suggest that patches of open prairie may have been Post Oak Savannah, these paleoenvironmental quite common within these woodlands. localities should reflect the long term westward movement and retreat of this boundary as climatic Pollen influx values decrease significantly in the upper conditions oscillate between moister and drier 52 cm of the core although the upper pollen conditions, respectively. Given that the Little River assemblages do not appear to change significantly in project area is in this same ecotonal setting, composition. Holloway et al. (1987:76) interpret this the paleoenvironmental reconstructions from these pattern as representing a significant change in the localities are highly relevant to modeling changes in structure of the vegetation community in the vicinity resource structure and base in the vicinity of the of the bog. Specifically, the lower influx value seem sites in question. to suggest a more open savannah woodland setting with a greater herbaceous understory and a more Given that all indications are that the two sites prevalent grassland component. According to Holloway were occupied during the late Holocene, the et al. (1987:76–77), this change appears to reflect a shift paleoenvironmental review will focus primarily on this from more mesic to drier conditions and may represent 3,000-year period. Although a number of bog locations the establishment of the Post Oak Savannah along this are present within the general vicinity of the Little portion of the deciduous forest. This change appears to River project, many of them (e.g., Gause Bog) do not have occurred sometime between 2125 and 1550 BP. contain a pollen record from the portion of the post- This date is younger than the 3000 Years BP date glacial period. Similarly, the 3700–3850 BP date from suggested by Bryant based on his analysis of the pollen 40–50 cm bs at Boriack Bog (Bryant 1977:146) record from Gause and Boriack Bogs (1977). Bryant suggests that the pollen record from the last nearly (1977) concluded that the Oak Savannah vegetation belt four thousand years of deposition is missing. However, within which the project area lies was established no Weakly Bog contains a record of this approximately later than 3,000 years ago and that the vegetation 3,000-year portion of the late Holocene. Here a date community has changed little since then. of 2375 radiocarbon years BP was obtained for a sample from 120–129 cm bs, and a date of 2125 BP was Interestingly, as pointed out by Bousman (1998:207), obtained for a zone between 58–65 cm bs. A date of the relative frequencies of grass and compositeae 1550 BP obtained from 50–58 cm bs suggests a pollen does not increase as would be expected 500–600-year hiatus in sedimentation in the core at a throughout the upper portion of the column. Rather, depth of about 55 cm bs. The upper 20–30 cm bs only two well-defined and isolated peaks in grass and represent essentially modern deposits (Holloway et compositeae pollen are evident. One occurs sometime al. 1987:Table 1), and using a linear regression formula around 1150 BP at a depth of about 43–53 cm bs, while between radiocarbon sample depth and age, Bousman the other, based on Bousman’s (1998:206–207) (1998:207) suggests that the upper 16 cm of deposits regression formula, is estimated to have occurred should date after ca. 400–500 BP. sometime around 400–500 years BP at a depth of between 10–20 cm bs. These spikes in grass pollen Holloway et al. (1987:75–77) interpret the lower coincide with decreases in oak pollen and might reflect section of the pollen spectra (58–124 cm bs) as short periods of perhaps warmer and drier conditions representing a woodland plant community that appears resulting in the expansion of the prairie patches or the to have been denser than that covering the vicinity of wholesale eastward retreat of the oak savannah. the bog today. In part, this interpretation is based on Bousman (1998:207) also suggests that the presence the extremely high pollen influx values for all taxa of pecan/hickory throughout to upper 75 cm portion including grasses, in conjunction with the rapid of the column, and the appearance of pine pollen in sedimentation rates characterizing these deposits. the upper 30 cm, in conjunction with the steady Arboreal pollen peaks between 75–104 cm bs starting increase in arboreal pollen –with the exception of the immediately below a 2–3 cm thick grass mat layer two grass spikes– may actually be interpreted as that coincides with a decrease in both Poaceae and indicative of increased moisture levels though the Late Cyperaceae pollen. The high influx values for grasses Holocene, contrary to Holloway et al. (1987).

9 Due to their relative proximity to the project area, three Sites 41MM340 and 41MM341 are located within the of the more useful paleoenvironmental localities are floodplain of the Little River, a setting with high water Gause, Patschke and Franklin bogs. Unfortunately, tables and therefore higher effective moisture levels although Patschke bog yielded a 22 foot column, than the adjacent upland. Such settings would have Gause produced a 14 foot column, and Franklin Bog always favored the development of woodlands and had a 10 foot column, no dates were obtained for any forests composed of high-moisture loving and/or portion of these columns. In the Gause column tolerant species. It is likely that shifts between graminea pollen dominated, ranging between 30 to vegetation communities dominated by woodlands and 70 percent of the pollen. The deeper levels (below grasslands would have occurred primarily under 9 feet) contain between 10–30 percent alder and relatively severe and long term drought conditions. chestnut pollen and chestnut continues to be present Perhaps a more accurate picture of the effects of these in small percentages through the 4 foot level. Oak paleoenvironmental conditions on the immediate pollen is present in small percentages (1–10%) below surroundings of the site is to view them as resulting in the 10 foot levels but increases in proportion thereaf- oscillations between more closed to more open ter with one exception, a decrease in the 5–6 and 6–7 woodlands characterized by heavy woody plant foot levels when grass pollen seems to spike. Although components in high moisture periods and a mosaic of grass and oak pollen seems to co-dominate the Franklin grasslands and woodlands during periods of lowered Bog column, pine and birch occur in small propor- water tables. The surrounding uplands, where drops tions throughout the upper 5 foot levels. The invasion in water tables would have affected vegetation of the oaks and grasses in the upper levels is common communities more drastically, are expected to have to both bogs and seems to be indicative of oscillated between relatively closed woodlands, more macroregional vegetation patterns. The occurrence of open woodland communities with patches birch and pine at Franklin bog may represent of grasslands or prairies and finally grasslands microregional differences compared to Gause bog. In as effective moisture levels decreased and water describing the pollen records from the two bogs the tables dropped. authors note the gross similarity between these profiles and that yielded by Patschke bog (Potzger and Tharp 1954:464). The only main difference between the latter and the former is the more pronounced alder-chestnut period in the Patschke bog record compared to the other two bogs.

Using Bousman’s (1998) reconstruction of estimated arboreal canopy cover extrapolating from the upper portion of Boriack Bog and using the data from Weakly Bog, the 2500 year period represented in the four occupation zones from 41MM340 has been characterized by decreasing arboreal cover between 3200-3000 BP reaching a relatively open grassland community between 3000 and 2000 BP. The increase in arboreal canopy cover between 2000–1600 BP signaled the return of woodlands and a more substantial woody vegetation component. A vegetation community dominated by grasses may have been present between 1000–1600 BP.

10 Chapter 4: Archaeological Background

Limited professional research has concerned Little Conversely, the remainder of the county, as well as River basin proper within Milam County. With the surrounding counties, have undergone extensive exception of a single project, the investigations archaeological investigations due, primarily, to the regarding portions of Little River have all dealt with presence of numerous surface lignite mines in the roadway, utility, or pipeline easements at the stream region. Most proximal is the Sandow Mine, which crossing. The single exception, an investigation of a extends from the south-central portion of Milam small area of stream-bank erosion immediately County, southwest through the northwestern portion surrounding the City of Cameron water pumping of Lee County; the subject of numerous investigations station, encountered no cultural material (Fickel 1993). locating abundant prehistoric sites (e.g., Rogers and TxDOT recorded two sites (41MM289 and Kotter 1995; Rogers 1999). 41MM344) in addition to the subject sites of this report, during survey efforts for roadway improvement (Ahr and Abbott 1999; TSDHPT 1990; TxDOT 1999). Previously Recorded Sites Espey, Huston, & Associates recorded a single site (41MM136) during a survey for Brazos Electric (Texas At the time of the current investigations, there were Archeological Sites Atlas). Survey for the All- 17 previously recorded archaeological sites along American Pipeline project, sponsored by the Bureau Little River in Milam County. Of these, only two sites of Land Management, lead to the discovery of two (41MM273 and 41MM289) consist solely of a sites (41MM272 and 41MM273) near the crossing at historic component (Turpin et al. 1992; TSDHPT Little River (Turpin et al. 1992). 1990). An inventory of the prehistoric sites is presented below in tabular form (Table 4-1).

Table 4-1. Previously recorded Prehistoric Archaeological Sites along Little River in Milam County

SITESITE COMPONENT SOIL LANDFORM RELICT EXTANT SITESITE SIZE QUAD PROJECT 41MM8 unknown ? terrace 20 740 ? Hanover Freeman Site 41MM12 unknown sandy loam gravel pit 60 2020 10acres Hanover CK Chandler 41MM13 unknown sandy loam terrace 70 1500 800x200m Hanover CK Chandler, Espey, & SMU 41MM14 unknown sandy loam floodplain 40 1420 20acres Hanover CK Chandler; SMU 41MM36 ? gravel terrace * <20 ? Little River Frank Bryan’s #136 (1930s) 41MM128 unknown sandy loam upland # 1560 ? Gause Bill Moore (T exas A&M) 41MM129 unknown sandy upland # 1860 ? Gause Bill Moore (T exas A&M) 41MM130 Late Prehistoric sandy terrace * 1380 ? Gause Bill Moore (T exas A&M) 41MM131 unknown sandy terrace * 820 ? Gause Bill Moore (T exas A&M) 41MM136 unknown sandy loam floodplain * 140 12.5acres Hanover EHA, Brazos Electric 41MM142 unknown ? terrace * <20 ? Sharp D Crawford 41MM253 ? ? upland # 610 ? Gause ? 41MM264 ? ? floodplain 20 1400 ? Pettibone ? 41MM292 Late Prehistoric ? floodplain * <20 50x100ft Hanover Texas A&M 41MM302 unknown sandy loam upland # 250 100x200 Gause Bill Moore (BVRA) 41MM344 Late Arch aic gravel terrace 20 280 ? Pettibone TxDOT FM 1600

avg 40 873.75

* - position relative to floodplain suggests adjacency with relict meander (<20) # - located atop upland bordering floodplain ? - data not contained in report or site form

11 Little River enters Milam County from the west at the border of Bell County, follows a sinuous course, trending east-southeast through the county, toward the eastern county line, where it confluences with the Brazos River. Throughout, the Little River maintains an exceptionally wide floodplain; over 5 km (3 mi.) wide at certain locales. Comparison of aerial photographs and USGS topographic quadrangles reveals the history of a once dynamic stream dissecting the floodplain with numerous relict scars of former meanders.

Eight of the sites appear to be located atop terrace formations along the border of the floodplain. Four of the sites occur atop the upland, ephemerally associated with Little River proper. These upland sites appear to be exploiting the gravel deposits common along the Little River in this portion of Milam County in this location. The remaining four sites occur within the floodplain, probably representing brief encampments on former minor terraces no longer visible on the current landscape. Calculating the distance of sites to the extant meander of Little River, it is interesting to note that the average distance is approximately 870 m (~½ mile); certainly a substantial distance from such a frequently utilized resource. As such, the probable location relative to the associated relict meander has been calculated through examination of the aerial photographs and USGS maps.

Table 4-1 shows the results of this likely reconstruction, with an average, corrected distance of sites to relict meanders of Little River approximating 40 m (130 ft.). This corrected distance indicates a more feasible proximity of prehistoric sites to the former meander of the stream. Within such a wide, dynamic floodplain, the distance of 40 m allows for the variability of a changing stream and, more importantly, suggests former landforms of preferred occupation.

12 Chapter 5: Geoarchaeology

Lee C. Nordt

The purpose of the preliminary geoarchaeological Pleistocene. The floodplain soils are mapped primarily report is to present the stratigraphic framework for as occasionally flooded, clayey Mollisols (Frio series) alluvial deposits and paleosols across the Little River with weakly developed and calcareous profiles. floodplain, and to make general inferences about prehistoric site formation processes and preservation Several sloughs are present across the floodplain in potentials. The numerical alluvial chronology is based the study area. Cutbank exposures from the Little River on a combination of radiocarbon ages from the current to the west of the project area display clayey, channel testing project (CAR) and the previous survey project fills that may correlate with the sloughs in the project (TxDOT). Detailed soil-stratigraphic descriptions are area. Adjacent to the sloughs, the floodplain given in Appendix A. elevation is 0.5–1 m higher.

Geology and Geomorphology Alluvial Stratigraphy The Little River begins at the confluence of the Leon Valley Cross-Section and Lampasas rivers approximately 60 km upstream An alluvial stratigraphic cross-section of the Little from the project area. Here the tributary network flows River valley in the project area was constructed using through upper Cretaceous formations consisting of a combination of the following sources: examination limestones and calcareous muds. In the project area, of sediment cores described during bridge construction the Little River flows along the strike separating the in the 1930s (Figures 5-1 and 5-2), investigation of Tertiary Midway Group to the north and the Tertiary cut bank sections of the Little River upstream from Wilcox Group to the south. The associated formations the project area, and study of re-excavated and/or principally contain inter-bedded sandstones and newly excavated backhoe trenches across the mudstones. The Little River confluences with floodplain in the vicinity of the two sites. Therefore, the Brazos River about 50 km to the east of the the model presented is somewhat different from project area. that originally developed by TxDOT to explain valley stratigraphy. The Little River is a highly sinuous, suspended-load stream with a floodplain width of about 1 km in the The surface topography from the recent archaeological project area (see Figure 1-1, 3-1). The modern channel survey and from field observations, documented the is flowing on alluvium and is entrenched by 6–7 m location of three drainage sloughs (C-7, C-8; C-10, relative to the adjacent floodplain surface. Previous C-11; C-17), two ridges (C-2, C-12), and intervening hydrological analysis from a local water-discharge floodplain areas. From the cross-section, three strath gauging station indicates that the floodplain surfaces cut into bedrock were identified. In the is inundated with floodwaters approximately once vicinity of C-1 through C-4 the high strath is overlain every two years. by 6–7 m of alluvium. The intermediate strath, from C-5 to C-17, was cut approximately 3 m deeper into Pleistocene alluvial terraces are mapped on both sides bedrock, with the overlying alluvium ranging from of the Little River floodplain adjacent to the project 9–11-m thick. The third strath was cut by the modern area. Unpublished soil survey maps delineate these channel and is another 3 m deeper relative to the terraces mainly as Alfisols with decalcified A-Bt intermediate strath (C-18 through C-20). The overlying profiles (Minwell series), consistent with a late alluvium is up to 13-m thick.

13 Figure 5-1. Little River basin, Milam County, Texas.

14 Figure 5-2. Alluvial stratigraphic cross-section on the Project Area. Based on deep sediment cores from bridge borings taken in the 1930s. Also shown, backhoe and cutbank localities and the location of extant drainage sloughs.

Three alluvial facies were identified from the sediment same channel. Further evidence for this scenario are cores: gravel-sand (channel), sand (point bar), and clay the presence of well developed paleosol sequences in (floodbasin). Above the high strath on the north end the floodbasin facies associated with the ridges above of the valley, channel and point bar facies are overlain both the high and intermediate straths (C-2 and by about 4 m of floodbasin clays. One of the surface C-12). In river cutbanks west of the project area, ridges occurs in this area (C-2). Above the intermediate multiple gray to grayish brown channel fills were strath there are three areas where thick floodbasin clays observed cut into yellowish brown and loamy point bury thin channel gravels or point bar sands (C-6, bar deposits, and buried by modern overbank sands. C-7; C-9 through C-11; C-17). The intervening areas This indicates that the thick floodbasin clays observed consist of much thicker channel and point bar facies, in the deep cores unconformably overlie the channel with the second ridge occurring over one of these areas and point bar facies and were deposited by a high (C-12). Thick channel, point bar, and channel margin sinuosity stream with multiple channel cut-offs. facies are associated with the modern channel in Ultimately the Little River valley filled to a common C-18 through C-20. elevation with continued floodbasin deposition from the modern channel. The upper elevation of the channel and point bar facies correlates across the high and intermediate straths Backhoe Trenches indicating that the same channel filled the valley in Fourteen backhoe trenches were opened across the this area (compare C-1 through C-4 to C-12 through floodplain within the project area. These trenches were C-16). It follows that the underling straths were reopened from the previous archaeological survey scarified to different depths across the valley by the project and described again stratigraphically and

15 pedologically for internal consistency. Several cutbank Unit 2 exposures from the Little River were also Unit 2 stratigraphically overlies Unit 1. Unit 2 probably schematically interpreted. Three unconformably forms the upper part of the thick clayey channel facies bound stratigraphic units were identified from associated with the modern sloughs in the valley cross- exposures provided by the backhoe trenches and section and in river cutbanks. In addition, Unit 2 forms cutbank exposures. These units are labeled Unit 1, a floodbasin veneer outside of the sloughs. In, or Unit 2, and Unit 3, from oldest to youngest. immediately adjacent to the modern sloughs, Unit 2 occurs in BHT-16, BHT-18, BHT-7, BHT-12, Unit 1 BHT-13, and BHT-5 (see Figure 5-3). In floodbasin Unit 1 is buried beneath the floodplain ridges in the positions where burying Unit 1, Unit 2 occurs in vicinity of BHT-A, BHT-B, BHT-8, BHT-26, BHT-A, BHT-B, BHT-8, BHT-26, BHT-15, and BHT-15, and BHT-24 (Figure 5-3; Appendix A). From BHT-24 (see Figure 5-3). In the sloughs, Unit 2 the valley cross-section and from cutbank exposures, contains faint bedding planes in the lower part, but is Unit 1 in the project area is probably the floodbasin weathered to a weakly developed paleosol at the top. facies of the thicker channel and point bar facies The paleosol is sometimes split by flood units. The identified above the high and intermediate straths (see Unit 2 paleosol consists of weak A-C or A-Bw/Bk Figure 5-3). A period of landscape stability and soil profiles. The surface horizon is dark gray to dark formation terminated deposition of Unit 1. The grayish brown and the subsurface typically grayish subsequent clayey paleosol developed an A-Bk profile. brown to gray. Iron concentrations in the lower part The surface horizons are very dark gray to black and indicate that a fluctuating water table and anaerobic the Bk horizons grayish brown, brown, and olive conditions existed during formation of the paleosol in brown with blocky to prismatic structure. The Bk the slough. horizons contain 5–10 percent carbonate filaments and in some cases 2–3 percent carbonate nodules. The In the floodbasin setting, the Unit 2 paleosol is less thickness of this paleosol is at least 2 m, and in the stratified because it accumulated more slowly and has lower part, soft iron masses indicate that at some time consequently been affected more by pedogenesis. in the past a fluctuating water table produced anaerobic The paleosol typically has clayey A-Bw horizon conditions in this zone. sequences. The surface horizons are very dark to dark gray and the subsoils dark gray to dark grayish Seven radiocarbon ages were obtained from within brown. A few profiles have up to two percent carbonate Unit 1 and four from features resting on top of the filaments. Unit 1 paleosol. The oldest age obtained from recent dating (CAR) within Unit 1 sediments is 3070±40 BP Radiocarbon ages from within Unit 2 range from and the youngest 2760±30 BP (see Figure 5-3). 2600±50 to 1500±70 BP (see Figure 5-3). Radiocarbon However, a bulk humate age of 4720±70 BP was ages from hearths resting on top of the paleosol of determined in the previous survey investigation Unit 2 range from 940±40 to 1270±40 BP (see (TxDOT) from a depth of nearly 3 m, indicating that Figure 5-3). These ages indicate that deposition was deposition of Unit 1 was underway by approximately underway possibly as early as 2600 BP and proceeded this time (see Figure 5-3). The ages bracketing features until no later than 940 BP. The interval of pedogenesis resting on top of the Unit 1 paleosol range from that occurred between the termination of deposition 2420±40 to 2600±50 BP (see Figure 5-3). Thus, of Unit 2 and the beginning of deposition of Unit 3 is deposition of Unit 1 could have terminated as early as unknown other than to say the maximum could have 2600 BP, but no later than 2420 BP. Based on been no more than a few hundred years. It is likely radiocarbon ages from the overlying stratigraphic unit that the interval was even less, and that the paleosol that approximate those of Unit 1, the paleosol in in the upper part of Unit 2 is cumulic, having formed Unit 1 is probably cumulic, having formed during a during a period of decreased flood deposition. period of depositional slowdown, and not during a period of complete cessation of flood sedimentation.

16 Figure 5-3. Detailed soil-stratigraphic profiles from backhoe localities showing the alluvial stratigraphic units, paleosols, radiocarbon ages, and location of archaeological materials.

17 Unit 3 2500 BP. The surface of Unit 1 has been altered to a The channel facies of Unit 3 is deeply buried in C-18 well-developed paleosol containing filaments and through C-20 adjacent to the Little River (see nodules of calcium carbonate. The interval of Figure 5-2). Point bar and channel margin facies of pedogenesis probably included the later stages the modern channel show thick stratified sands and of Unit 1 deposition when flood frequencies silts as revealed in CB-6 and BHT-2 (see Figures 5-2 were decreasing. and 5-3). The lateral unconformity separating the Unit 3 channel deposits and older floodplain deposits Unit 2 stratigraphically overlies Unit 1 (see of Unit 1 and Unit 2 probably occurs between C-17 Figure 5-4). Unit 2 consists of multiple abandoned and C-18 (see Figure 5-2). Colors are typically dark channels inset to Unit 1, and also of a thin overbank gray, brown, and pale brown in the channel facies. facies burying Unit 1 in floodbasin positions. The Unit 3 deposits also fill the upper part of the sloughs channel facies is dominated by silts and clays with a that cross the floodplain in the project area as thin basal sand layer, and ranges from a few to 6 m illustrated in BHT-16, BHT-18, BHT-7, and BHT-5 thick. Facies relationships indicate deposition by a (see Figures 5-2 and 5-3). These deposits are meandering, suspended-load stream. Deposition of characterized by clayey and occasionally stratified Unit 2 occurred between approximately 2500 and loams. A third location for Unit 3 occurs in the 900 BP. The upper part has been pedogenically altered floodbasin setting where it buries Unit 2 in BHT-A, to a weakly expressed paleosol. This indicates that BHT-B, BHT-8, BHT-26, BHT-12, BHT-13, BHT-15, only a brief period of depositional slowdown occurred and BHT-24 (see Figure 5-3). Unit 3 in these settings between the shift from deposition of Unit 2 to Unit 3. reveals similar colors, textures, and degree of Paleo-sloughs of Unit 2 appear to be generally related stratification as in slough positions. Unit 3 typically to the position of the modern sloughs. contains an abundance of unweathered shale clasts, particularly in fine-grained units. Periodically, very A period of bedrock incision occurred with the brief intervals of depositional slowdown and initiation of deposition of Unit 3 (see Figure 5-4). pedogenesis occurred during deposition of Unit 3. Unit 3 is up to 13-m thick near the modern river where it is dominated by a thick channel facies. Modern Based on radiocarbon ages from Unit 2, Unit 3 must sloughs are also partially filled with Unit 3 deposits be younger than 940 BP. However, radiocarbon because they periodically carry high-energy ages directly from Unit 3 range from 1050±50 to floodwaters as evidenced by basal sands and silts. The 420±40 BP. The oldest age was obtained from bulk floodbasin facies of Unit 3 forms the modern humates (TxDOT) and may be dating somewhat older floodplain surface across the project area. Deposition than the time of deposition of the associated sediments. of Unit 3 has occurred during the last 900 years. Unit 3 sediments are accumulating today as they are associated with flooding from the modern river. Geoarchaeological Interpretations

Stratigraphic Overview Prehistoric preservation potentials and site formation processes were influenced by geological and Figure 5-4 displays a schematic alluvial stratigraphic pedological processes in the study area. Unit 1 is either cross-section of the project area. Unit 1 is the oldest buried by 50–150 cm of floodbasin clays or eroded unit and ranges in thickness from 6–11 m. The basal by Unit 2 channels to depths of 6–7 m (see contact unconformably rests on a bedrock strath. Figure 5-3). As the Unit 1 channel migrated across Unit 1 consists of a thick channel and floodbasin fa- the valley, the floodbasin facies probably accumulated cies spanning the entire width of the floodplain. Fa- in a relatively slowly aggrading environment. Using cies relationships suggest deposition by a meandering, modern hydrological conditions of high flood mixed-load stream. Deposition of Unit 1 began no later frequency and duration, relatively rapid deposition of than 4700 BP and proceeded until approximately clays could have occurred in the past. The presence

18 Figure 5-4. Schematic alluvial stratigraphic cross-section of the Project Area.

of two distinct occupation zones within the Bk horizon and subjected to cracking upon desiccation. These of the Unit 1 paleosol supports this interpretation (see factors could serve as vectors for mixing Figure 5-3). If deposition had been slow, an archaeological materials. The paleo-location of the over-thickened A horizon would extend to the depth Little River channel and associated sloughs at the time of the occupation zones. Occupation zones also occur of Unit 1 deposition is difficult to decipher. Following on top of the Unit 1 paleosol. Based on radiocarbon the geological model presented above, the sloughs as ages this paleosol is cumulic, having formed during identified today and associated with Unit 2 probably slow deposition such that pedogenesis kept pace with did not exist during Unit 1 times. deposition rather than forming during a discrete interval of non-deposition. Although high magnitude Preservation potentials and site formation processes floods periodically occur in the project area, the are similar for Unit 2 where it buries Unit 1 in floodwaters in prehistoric times were probably not floodbasin settings (see Figure 5-3). The difference is such that significant erosion occurred outside of the that the interval of depositional slowdown at the end sloughs. Based on environment of deposition, of Unit 2 times was contracted based on minimal preservation potentials for discrete occupation zones pedogenesis in the associated paleosol. However, with site integrity is high within the floodbasin facies where Unit 2 fills sloughs, archaeological of Unit 1 and probably on top of the Unit 1 paleosol interpretations will differ. Here, deposition was as well. According to radiocarbon ages, the floodbasin probably more rapid as denoted by distinct silt and facies of Unit 1 would contain buried sites dating sand bedding planes at depth. This would enhance the to the Middle Archaic and to the early part of the preservation of discrete occupation zones, although Late Archaic. subsequent high magnitude floods could have occasionally destroyed previous occupation surfaces. Mitigating contextual integrity for Unit 1, however, Prehistoric occupation zones in Unit 2 would date to are post-depositional pedogenic processes. Numerous the latter part of the Late Archaic and to the first part biocasts were identified in the Unit 1 paleosol as a of the Late Prehistoric. Overall, preservation potentials result of burrowing organisms. Although the paleosol in Unit 2 should be considered at least moderate given contains very few slickensides, it is nevertheless clayey the dominance of fine-grained textures and deposition

19 in a relatively low energy environment. Cracking clays and bioturbation are also factors mitigating preservation potentials in a primary context. However, the sloughs probably do not desiccate as frequently as the floodbasin settings.

Based on the appearance of continued migration to the southeastern part of the valley, the Little River was probably closer to sites 41MM340 and 41MM341 during their formation. It is unclear whether the Little River actually occupied one of the sloughs in the project area during formation of the two sites. Even if it did not, based on the presence of a channel facies and on the thickness of Unit 2, it appears that the sloughs were carrying floodwaters more frequently during occupation than today. Thus, prehistoric occupation in floodbasin settings adjacent to the sloughs in Unit 2 probably occurred closer to a water source than what would be apparent by observing geomorphic conditions today.

Based on the absence of paleosols and presence of stratified coarse textures, deposition of Unit 3 in the project area probably occurred rapidly. Rapid deposition would lead to better preservation of discrete occupation zones, but frequent, and perhaps high velocity floods, would tend to rework features into secondary contexts. Any discovered archaeologi- cal materials would date to the Late Prehistoric or Historic.

20 Chapter 6: Methodology

Pursuant to the contract between TxDOT and CAR, a define the nature of these deposits and aid their manual total of five 2-m2 blocks were excavated to investi- excavation, two backhoe trenches were excavated gate the two sites. The location of three of these blocks within the site limits connecting two separate pairs of was prescribed under the provisions of the contract. auger borings located in the southern portion of The placement of the remaining two blocks was to be 41MM340. Each trench was excavated to approxi- determined in field by means of exploratory methods. mately 1.6 m (5.25 ft.), the terminal excavation depth The prescribed blocks were intended to investigate as prescribed by TxDOT. specific features exposed in backhoe trenches during the survey phase. The additional blocks were to ex- Prior to manual excavation of each block, the upper plore areas of rich cultural deposits. 30 cm (at site 41MM340) and 40 cm (at site 41MM341) of sediment from ¾ of each block (three 1-m2 test units) was mechanically removed. Three of Mechanical Excavations the four 1-m2 test units in each block were thus exca- vated from 30 cm or 40 cm to 160 cm below ground In an attempt to relocate features designated for exca- surface. The remaining test unit within each block was vation and prospect for comparative stratigraphy, manually excavated from the original ground surface 14 backhoe trenches were reopened for the current to a depth of 1.6 m (5.25 ft.). In total, 16.5 m3 (582.6 investigations. Three trenches were reopened prima- ft.3) was excavated manually at 41MM340, and rily to investigate features previously encountered 10.4 m3 (367.2 ft.3) was excavated manually at during the survey phase. The remaining trenches were 41MM341. reopened for geomorphological interpretation of the ® floodplain. The trenches were excavated with a Case A settling pond was excavated adjacent to each water ® 580K backhoe equipped with an Extend-A-Hoe arm screening station to prevent particulate-laden water attachment. from entering nearby wetlands or Little River. The settling ponds measured approximately 10-m wide, To aid the identification of clusters or concentrations 20-m long, and 2-m deep. On site 41MM340 the pond of cultural deposits that would condition the place- was excavated off-site immediately south of the site ment of the additional blocks and influence future data boundary. On site 41MM341 the pond was dug off- recovery plans, a total of 202 auger borings were ex- site immediately north of the site boundary. cavated on the two sites. A five-meter grid was main- tained across the previously delimited site boundaries In order to conform to Occupational Safety and Health as identified by TxDOT during the survey phase. A standard farm tractor equipped with power takeoff Administration (OSHA) standards for protection of (PTO) and fitted with an auger attachment (9"Ø; 4' employees in excavations (29CFR1926.652), safety long) aptly excavated the borings with minimal benches were mechanically excavated around the difficulty. perimeter of settling ponds, backhoe trenches, and excavation blocks. Each safety bench was excavated Two areas of dense cultural materials with some evi- to adequate depths to maintain proper slope angles dence of stratification were identified along the south- required under these provisions. No significant depos- ern edge of site 41MM340 (see Manual Excavations). its or features were encountered during excavation of These areas appeared to be ideal locations for the two the safety benches. additional block excavations. However, to more clearly

21 Manual Excavations successfully utilized a method for screening clay ma- trices in the coastal southern California area using Based upon TxDOT specifications and auger boring sodium bicarbonate as a catalyst for dissolution of results, five 2-m2 excavation blocks were established excavated clay. More recently, the compound was uti- to investigate potentially significant deposits and fea- lized along coastal Texas processing dense silty clay tures. Three 2-m2 blocks were excavated at site alluvium (Mahoney and Moore 1998). As the sedi- 41MM340 within the existing right-of-way (Figure 6- ments and soils encountered within the Little River 1). One of these blocks, Block 1, was placed to inves- floodplain differed only slightly from those encoun- tigate deposits associated with features exposed in tered in coastal regions, it seemed likely that the Backhoe Trench 26 during survey work. The remain- compound should perform comparably. ing two block locations, Blocks 2 and 3, were placed to investigate apparent stratified cultural deposits as Chemically, the addition of sodium bicarbonate causes indicated during auger boring excavation. Each block an ion exchange reaction wherein sodium ions from was subdivided into four separate 1-m2 excavation the additive replace existing ions within the clay. This units and proveniences were maintained accordingly. reaction causes the aggregate bond of the clay to fail A total of 12, 1-m2 excavation units were dug. and thereby disperses the clay into individual particles. These deflocculated particles remain separated due Two 2-m2 blocks were excavated at site 41MM341, to their resultant charge. Aside from contributing an with one inside the current right-of-way and one within additional sodium component to the resultant the proposed right-of-way (Figure 6-2). Block 1 was processed matrix, the use of sodium bicarbonate placed to investigate deposits associated with features should have no adverse effect upon the ecology of the exposed in Backhoe Trench 24 during survey work. surrounding environment (Dennis Neilsen, TxDOT Block 2 was placed to investigate deposits associated Environmental Quality Specialist, personal commu- with features exposed in Backhoe Trench 15 during nication, 2000). survey. Approximately 1 lb. (454 g) of sodium bicarbonate All of the manually excavated soils and sediments was added to roughly 3.5 gal. (13 l) of excavated ma- were water screened through ¼" hardware cloth. Due terial, along with adequate water to just cover the mix- to the cohesive properties of the clays within this re- ture. This was allowed to soak in a five-gallon bucket gion, it was necessary to use a deflocculent agent to for at least an hour prior to water screening. During expedite processing of the soils. The addition of vari- the soaking process, the mixture was stirred once or ous compounds to accelerate processing of clayey soils twice to expose a greater surface area to the aqueous is common in archaeology. Laundry detergents, mixture. The resultant deflocculated material flowed Calgon®, and other forms of cleaning detergents have through the ¼" screens much more easily than all been utilized with less than satisfactory results otherwise would have been possible, saving (Mahoney and Moore 1998). Each of these compounds countless man-hours in the process. are classified as polyelectrolytes that, under certain conditions, act as an expedient method of A standard 1 m2, 10 cm-level of blocky clayey soils deflocculation. However, as the overall process of typically yields 12 five-gallon buckets filled ¾ full of ion replacement with clays is not completely excavated soils or sediments. Utilizing a formula understood (Grim 1968:212), various geochemical where 1 lb. (454 g) of sodium bicarbonate is used per factors must be calculated for the proper formula for five-gallon bucket, each 10 cm-level would require each specific clay. 12 lb. (5.4 kg) of the compound to efficiently process the excavated material. Applying the formula to the As such, a trial-by-error approach was used to deter- current project, approximately 3,228 lb. (1467 kg) was mine the most efficient means by which to deflocculate required to deflocculate all of the manually excavated the soils for water screening. Van Horn, et al. (1993) soils and sediments.

22 Figure 6-1. EDM-derived topographic map of 41MM340.

23 Figure 6-2. EDM-derived topographic map of 41MM341.

24 Of concern to previous projects utilizing this to the cultural zones encountered. More specifically, deflocculation technique was the effect of the com- each zone was defined as a unique provenience, fur- pound additive upon certain artifacts for dating pur- ther delineated by 10-cm arbitrary levels within each poses. Specifically, the effect of the compound upon provenience. In a number of cases, the arbitrary level charcoal has not previously been addressed in an ar- within a zone was less than 10-cm in thickness. chaeological context (Van Horn et al. 1993; Mahoney and Moore 1998). In consultation with Ron Hatfield, Various samples were collected in field to provide Beta Analytic, Inc., Laboratory Manager, it was relevant data for determining integrity and eligibility. determined that processing materials with sodium A soil sample was collected from each arbitrary or bicarbonate will not effect the absolute dating of char- cultural stratigraphic level of one unit within each coal specimens as long as the laboratory is notified, 2-m2 block at sites 41MM340 and 41MM341 through- prior to processing, that sodium bicarbonate was used out the vertical column. Macrobotanical (flotation as the deflocculent (personal communication, 2000). processing) and microbotanical (e.g., pollen, diatoms, and phytoliths) samples were similarly collected. Water was supplied for the processing stations by a Mussel shell and 14C samples were collected as 2000-gallon diesel tanker truck. On average, approxi- encountered and maintained with provenience. mately 1,500 gallons of water was used per full day Furthermore, all of these samples also were collected of excavation. Water was sumped from the tanker via from each individual feature encountered during the a 5 hp engine with a water pump attachment via a 2½" manual excavations. inlet hose. Water was then diverted to each individual station by means of a ¾" typical garden hose with a All cultural material encountered during manual and cut-off spray nozzle attached. mechanical excavation was collected and recorded on unique field forms by provenience. These materials In order to facilitate the capture of runoff water, a were processed and analyzed in a controlled settling pond was excavated adjacent to each water laboratory setting; all recovered cultural materials are screen station to prevent particulate-laden water from curated at CAR. entering nearby wetlands or Little River. Further, silt- screen fencing was erected around the perimeter of each settling pond to capture potential runoff in the event of overflow or heavy rains.

Excavation Strategy

All of the blocks prescribed by TxDOT (Blocks 1 and 2 of 41MM341 and Block 1 of 41MM340) focused specifically upon individual features and exhibited no obvious stratigraphy dictating the necessity of strati- graphic excavation. As such, the test units excavated within these three blocks were dug in arbitrary 10-cm levels below ground surface.

However, Blocks 2 and 3 of 41MM340 did exhibit some evidence of stratified cultural deposits. As such, we attempted to excavate the eight units contained within these two blocks stratigraphically, according

25

Chapter 7: Results

Richard Mahoney, Steve Tomka, Raymond Mauldin

This chapter presents a detailed discussion of the re- increased southward for approximately 25 m (82 ft.), sults of the excavations at both 41MM340 and at which point another negative corridor was encoun- 41MM341. Supporting data for many of the sections tered (see Figure 6-1). Similarities in soil texture and are presented in the 10 appendices provided at the end moisture to the northern corridor suggest the possi- of this volume. bility of a second stream channel. If correctly inter- preted, the two ancient channels may have defined a narrow, elevated landform of preferred occupation. 41MM340 Cultural Stratigraphy Auger Data The stratigraphy revealed by the mechanical and Excavation of 81 auger borings at 41MM340 revealed manual excavations across the southern portion of the two distinct areal concentrations of cultural material site, represented by excavations in Blocks 2 and 3, within the site boundaries (see Figure 6-1). A 20-m was complex. In each of these blocks, we were able to (66-ft.) corridor of negative auger borings separates identify nine different zones. These zones, numbered the two concentrations. Each of the negative auger 1 (top) through 9, were defined by relatively subtle borings within this corridor exhibited a moist to very differences in sediment color and texture. These zones wet soil/sediment consistency throughout. These data also contain differences in artifact content. Figure 7-1 suggest that the two concentrations appear to have presents a photo of Unit C, Block 2, at 41MM340 with been formerly separated by a small slough or possibly the zones identified. Figures 7-2 and 7-3 present cor- an intermittent stream. It is equally possible, however, responding profiles for excavation units C and D in that the slough post-dates occupation of the site and Block 2 and units C and A in Block 3. has eroded through the cultural deposits. Note that not all zones were represented in all exca- The northern section of this site contained a relatively vation units. For example, reference to Figure 7-2 sparse amount of cultural material as evidenced by demonstrates that while in Unit C of Block 2, all nine auger borings and backhoe trench excavations. While zones were present, in Unit D, Zones 2 and 3 were a diverse assemblage of cultural material was encoun- truncated, being represented only in the western por- tered, the density thereof was relegated to a relatively tion of the profile. This can be seen in Figure 7-4, a thin, inconsistent zone of recovery located approxi- photo of all four units in Block 2. Zone 2 in this Block, mately 60 cm (24 in.) below ground surface. Sparse characterized by a shell lens, is clearly visible in three to moderate scattering of fire-cracked rock, fired sand- of the units, as well as the profile of the backhoe trench stone, chipped stone, mussel shell, bone, and char- at the top of the photo, but is not present in Unit B. coal comprise the majority of the recovered cultural The discontinuity in zones is also demonstrated by materials from this section. the observation that only Unit C within this Block contained all nine zones. A similar situation can be In contrast, the southern section of the site revealed a observed in Block 3, excavation unit C (see Figure 7- relatively abundant amount of similar materials, with 3), where Zones 4, 5, and 6 all converge. at least two stratified components discernible within the auger borings. The density of materials was no- As noted in the previous chapter, the testing strategy ticeably higher than the northern section beginning focused on isolating these various zones. While the immediately south of the negative corridor. Densities thickness and slope of the zones fluctuated within a

27 given excavation unit, the excavation followed, to the Figure 7-1. South wall of excavation Unit C, Block 2, degree possible, these fluctuations. In some cases, at 41MM340 with zones identified. mixing of material from two different zones may have occurred, but we are confident that this was not com- mon. Finally, note that the northern portion of 41MM340, represented by the excavations in Block 1, did not exhibit any clear stratification. This area was excavated in 10 cm arbitrary levels. However, as discussed below, artifact data suggest that within this block, at least two, and possibly three, different con- centrations of cultural material are represented.

Excavations in Block 1 at 41MM340 recovered 1,975 pieces of chipped stone debitage. Figure 7-5 (top) pre- sents the distribution of debitage by arbitrary 10 cm levels for all units in this block. Two, and possibly three modes are apparent in the figure, with one at level 7 and a second and possibly a third mode repre- sented by levels 10 through 12. Examination of the distribution of debitage by level from the four indi- vidual units, an example of which (Unit B) is pre- sented in Figure 7-5 (bottom) demonstrates these same modes.

Figure 7-6 (top) presents a similar graph for the 4,899 chipped stone items recovered in all units from Block 2. Here, the number of debitage has been standard- ized to account for the different volume of material excavated in each zone. Clearly, four different peaks are represented, with Zones 2, 4, 6, and 8 all having a high density of material. The bottom bar graph in Fig- ure 7-6 represents Unit A in this block. Note that Zone 3 is not reflected in this unit, but peaks in debitage in the four primary zones are clearly present. Figure 7-7 (left) presents standardized fire-cracked rock weights These comparisons demonstrate that the relationships for Block 2, while the right-hand graph in Figure 7-7 between the nine zones represented in each block are presents standardized weights for all mussel shell. complex. For example, it is not necessarily the case These patterns, while they vary by material, reinforce that Zone 4 in Block 2 corresponds to Zone 4 in Block the identification of Zones 2, 4, 6, and 8 as being dis- 3. Nevertheless, both Blocks 2 and 3, separated by tinctive areas with high concentrations of materials less than 10 meters, contained nine definable strati- separated by zones of lower material density. graphic zones. These zones had roughly similar sedi- ment structure, color, and texture. The zones differ in Finally, Figure 7-8 presents a bar graph of chipped detail in terms of chipped stone, fire-cracked rock, and stone recovered by zone in Block 3. As with the Block ecofactual content, but in each block there are data to 2 data, these numbers have been standardized to re- suggest that four distinctive, cultural strata can be iden- flect different volumes of excavated material. Four tified. These zones are separated by zones that, though peaks are again present, though the chipped stone peak not sterile, have relatively lower counts and weights in Zone 2, which dominated Block 2, is not as promi- of artifacts and ecofacts. nent in this block.

28 Figure 7-2. Profile of south wall of excavation units C and D, Block 2, 41MM340.

Figure 7-3. Profile of south wall of excavation units C and A, Block 3, 41MM340.

29 Figure 7-4. Excavation units A, B, C, and D, Block 2, 41MM340. Note continuation of shell lens in western profile of associated backhoe trench.

      V H OV  Q H R  Y  Ã= H LF Ã/  Q K  R S WL  UD D J Y  WL  D D F  WU [ 6 (                                  &KLSSHGÃ6WRQHÃ'HELWDJHÃSHUÃÃFXELFÃFP 1XPEHUÃRIÃ&KLSSHGÃ6WRQHÃ,WHPV

  

  OV  V H H Y Q H  R  Ã/ Ã= Q  LF LR W K  D  S Y UD D J F  WL [ D  (  WU 6  

  

              1XPEHUÃRIÃ&KLSSHGÃ6WRQHÃ,WHPV &KLSSHGÃ6WRQHÃ'HELWDJHÃSHUÃÃFXELFÃFP

Figure 7-5. Distribution of chipped stone by level for Figure 7-6. Distribution of chipped stone by level for all excavation units in Block 1 (top) and for only Unit all excavation units in Block 2 (top) and for only Unit B, Block 1 (bottom), 41MM340. A, Block 2 (bottom), 41MM340.

30  

 V  H Q R  V  Ã= H LF Q K R  S  Ã= UD LF J K WL  S  UD UD W J 6 WL   D WU 6  

 

                     )&5Ã:HLJKWÃ JU ÃSHUÃÃFXELFÃFP 0XVVHOÃ:HLJKWÃ JU ÃSHUÃÃFXELFÃFP

Figure 7-7. Standardized fire-cracked rock (left) and mussel shell (right) weights by zone, Block 2, 41MM340.

Beta Analytic, Inc., of Miami, Florida (Beta). All samples were corrected for differing fractionation, and

 all samples were calibrated using version 3.5 of OxCal

 (Ramsey 2000). The sample data and results are pro- V H  Q vided in Table 7-1. Appendix B presents the sample R Ã=  LF data sheets. K S  UD J WL  D WU Samples from Block 1 were recovered from the sur- 6 

 face and the base of the feature encountered during the

 survey phase within BHT 26. Previous Beta radiomet-

     ric dating for a charcoal sample recovered within this &KLSSHGÃ6WRQHÃ'HELWDJHÃSHUÃÃFXELFÃFP feature by TxDOT provided a corrected date of 2820±110 BP (Beta #128736). Current efforts have pro- Figure 7-8. Standardized volume of chipped stone duced dates of 2770±50 BP (Beta #141825) and 2880±40 recovered by zone in Block 3, 41MM340. BP (Beta #141826) from the surface and base of the feature, respectively. The feature, then, is securely dated Radiocarbon Dates by these three samples to about 970 B.C.

At the suggestion of TxDOT archaeologists, the sam- The sampling strategy of bracketing zones focused on pling strategy employed at 41MM340 attempted to the stratified cultural deposits in Blocks 2 and 3. A bracket each of the primary cultural zones encoun- total of 14 charcoal samples recovered from these two tered in Blocks 2 and 3 and date the single feature blocks were submitted for radiometric dating (Table encountered within Block 1. For Blocks 2 and 3, if we 7-1; Appendix B). For Block 2, carbon samples were assume that a given radiocarbon date provides an ac- extracted from the upper and lower portions of each curate date of the sediments in which it is deposited, cultural zone, with the exception of Zone 6, where dating samples from the upper and lower portion of appropriate samples were not encountered. All samples zones would provide a beginning and ending date for in Block 2 were taken from Unit C. Figure 7-9 super- that zone. imposes the vertical provenience of the samples on the profile for Block 2, Unit C. Note that in several A total of 16 charcoal samples from 41MM340, cho- cases, the samples represent pooled charcoal from a sen in consultation with TxDOT, were submitted to given level, while in others the vertical provenience

31 Table 7-1. Beta Analytic, Inc., results of radiocarbon dating for samples recovered from 41MM340

Calibrated BC Calibrated BC Corrected Range at 1 Range at 2 Block Zone Aspect Beta # Radiocarbon Age Sigma Sigma 1 n/a upper 141825 2770 +/- 50BP 980-830 1020-810 1 n/a lower 141826 2880 +/- 40BP 1130-990 1220-920 2 2 upper 141827 2380 +/- 50BP 760-390 770-370 2 2 lower 141828 2510 +/- 40BP 790-540 800-410 2 4 upper 141829 2600 +/- 50BP 830-590 900-540 2 4 lower 141830 2480 +/- 40BP 770-510 780-410 2 8 upper 141832 2760 +/- 30BP 970-830 1000-820 2 8 lower 141833 3070 +/- 40BP 1400-1260 1430-1210 3 2 upper 141834 2440 +/- 40BP 760-400 770-400 3 2 lower 141835 2420 +/- 40BP 760-400 770-390 3 4 upper 142625 2880 +/- 40BP 1130-990 1220-920 3 4 lower 142622 2950 +/- 40BP 1260-1050 1300-1010 3 6 upper 141838 2800 +/- 40BP 1000-900 1050-830 3 6 lower 142623 2910 +/- 40BP 1210-1010 1260-970 3 8 upper 142624 3060 +/- 50BP 1400-1260 1430-1130 3 8 lower 141840 3050 +/- 40BP 1390-1260 1410-1130 is restricted to a few centimeters. Though not shown, Using the 95.4 percent probability range of the upper the horizontal provenience of most samples was con- and lower dates within the blocks, the radiocarbon fined to small areas within Unit C. For Block 3, a to- dates suggest that the deposits form a relatively tight tal of 8 samples were submitted (Table 7-1). Unlike sequence. Assuming that the radiocarbon dates on the the Block 2 samples, where all radiocarbon dates came materials actually reflect the time span of the deposits from Unit C, the samples from Block 3 came from all in which the material was found, for Block 2 a mini- four units. Three samples were from Unit C, two from mum span of about 440 years (ca. 1210–770 B.C.) and Unit B, two from Unit A, and one from Unit D. While a maximum span of roughly 1,060 years (ca. 1430– this may complicate decisions regarding comparisons 370 B.C.) is suggested for the Zone 2 through Zone 8 between dates for a given zone, as well as any esti- deposits (Table 7-1). For Block 3, a minimum span of mate of differential sedimentation rates within the 360 years (ca. 1130–770 B.C.) and a maximum span of block, appropriate samples were not available from a 1,010 years (ca. 1410–400 B.C.) is suggested for the single unit within Block 3. Zone 2 through Zone 8 deposits (Table 7-1). Both blocks, then, represent deposits of the same general Examination of the calibrated and corrected ages at age, a relatively brief period that straddles the Late either the 1 or 2 sigma ranges will demonstrate that Archaic I and Late Archaic II Periods (Johnson most dates are in good stratigraphic order. However, 1995:73). there are two possible exceptions, one in each block. Beta sample numbers 141830, from the lower depos- In order to consider the ages of particular zones within its in Zone 4 of Block 2, and number 141838 taken a given block, clustering procedures suggested by from the upper deposits of Zone 6 in Block 3, may be Ward and Wilson (1978; Wilson and Ward 1981) were too recent when their overall stratigraphic relation- used. Initially, all radiocarbon dates within a block ships are considered. were tested using Ward and Wilson’s Case II assump- tions and the DSPLIT radiocarbon program (Kintigh

32 Reference to Figures 7-10 and 7-11 suggests that the dates from Zones 2 and 4, Block 2 reflect a similar time frame as Zone 2 in Block 3. Similarly, the lower date from Zone 8, Block 2, is similar to the two dates for Zone 8 in Block 3. While the relationship between Zones 4 and 6 in Block 3 and Zone 6 in Block 2 re- main unclear, it is likely that they also date to similar time frames. In spite of sampling complications, and the observation that in several cases the zones them- selves merge together or are truncated within each block (see Figures 7-2, 7-3), each block represents a similar range of time, and several of the zones within the blocks are temporally indistinguishable.

Table 7-2 presents additional information on radio- carbon dates and accumulation rates from Blocks 2 and 3. Here, we have used the interval estimate pro- vided by the calibration program to calculate the time between adjacent dates within a block. The procedure provides a probability estimate of the number of years represented between the two adjacent dates (Ramsey 2000), and both the 1 sigma and 2 sigma ranges are Figure 7-9. The vertical provenience of carbon provided in the table. Note that we have eliminated samples superimposed on the profile for Block 2, Unit the two potentially problematic dates in the table, C, 41MM340. though their inclusion does not dramatically change the overall pattern. Assuming, then, that the remain- ing radiocarbon dates reflect the time of sediment ac- 1992:83–85). For Block 2, DSPLIT produced three cumulation, and using the 2 sigma range as an example, groups of dates that were statistically indistinguish- the upper and lower dates for Zone 2, Block 2 suggest able. The first group consisted of the four upper dates, that there is a 95.4 percent probability that sediments representing Zones 2 and 4. The remaining two dates accumulated over a period of between 0 and 370 years. from Zone 8 were statistically different from this In Zone 8 of Block 2, the upper and lower dates sug- group, as well as statistically different from each other. gest that the accumulation took at least 280 years, and Figure 7-10 presents the probability distributions as not more than 570 years, at a 95.4 percent probability well as the 1 and 2 sigma ranges, with the three groups level. As it is equally probable that, for example, the identified. Zone 8, Block 2 accumulation occurred in 280 years as it is that the accumulation took 570 years, we also For the eight dates in Block 3, the DSPLIT procedure provide the midpoint of the 2 sigma range as a point also produced three groups. The first group contained estimate for the formation. Clearly, a number of other the two dates from Zone 2. The second group had four interpretations of these dates are possible, and Ap- dates, consisting of those in Zones 4 and 6, and the pendix B, as well as Tables 7-1 and 7-2, provide the third group contained both dates from Zone 8. Thus, reader with the appropriate data to explore alternative the two dates from Zone 2 are indistinguishable, as procedures. are the two dates from Zone 8. In this block, the middle zones, consisting of Zones 4, 5, and 6, appear to be Using the cumulative number of years in the midpoint indistinguishable. Figure 7-11 presents the probabil- estimate column in Table 7-2 suggests that Block 2 ity distributions for these dates, as well as the 1 and 2 formed in roughly 985 years, and Block 3 formed in sigma ranges, with the three groups identified. 1,275 years, estimates that are close to the maximum

33 Zones

2 U 2380±50BP

2L 2510±40BP

4 U 2600±50BP

4L 2480±40BP

8 U 2760±30BP

8L 3070±40BP

2000 1500 1000 500 0 Calibrated date BC Figure 7-10. Probability distributions of calibrated radiocarbon dates for Block 2, 41MM340. Groups, identified by color, produced by DSPLIT procedure. Atmospheric data from Stuiver et al. (1998); OxCal v3.5 Bronk Ramsey (2000); cub r:4 sd:12 prob usp[strat].

Zones

2 U 2440±40BP

2L 2420±40BP

4 U 2880±40BP

4L 2950±40BP

6 U 2800±40BP

6L 2910±40BP

8 U 3060±50BP

8L 3050±40BP

2000 1500 1000 500 0 Calibrated date BC Figure 7-11. Probability distributions of calibrated radiocarbon dates for Block 3, 41MM340. Groups, identified by color, produced by DSPLIT procedure. Atmospheric data from Stuiver et al. (1998); OxCal v3.5 Bronk Ramsey (2000); cub r:4 sd:12 prob usp[strat]. 34 Table 7-2. Radiocarbon interval estimates and deposition rates, 41MM340

14C date 14C date 1 sigma 2 sigma Midpoint of 2 Zones cm separating Accumulation Block Location 1 Location 2 interval range interval range sigma interval present dates rate (cm/100 years) 2 2u 2l 10-270 0-370 185 2 7 3.78 2 2l 4u 0-260 0-360 180 3 14 7.78 2 4u 8u 30-310 10-380 195 4,5,6,7 56 28.7 2 8u 8l 360-530 280-570 425 8 23 5.41 3 2u 2l 0-250 0-310 155 2 8 5.16 3 2l 4u 330-690 200-800 500 3 8 1.6 3 4u 4l 10-190 0-290 145 4 11 7.59 3 4l 6l 0-150 0-250 125 5,6 7 5.6 3 6l 8u 110-330 10-390 200 7 18 9 3 8u 8l 0-110 0-200 150 8 9 6 span suggested by comparing the upper and lower ra- years, and it could be the case that the accumulation diocarbon dates in each block. For example, we esti- was much faster. Nevertheless, the varying rates of mated that Block 2 minimally formed between ca. deposition, hinted at by the data presented in Table 7- 1210–770 B.C. (440 years) and 1430–370 B.C. (1,060 2, suggest that this rate was not constant. years). The cumulative number of years provided by the midpoint summary for this block, 985 years, is consistent with that estimate. Amino Acid Racemization Results

The final two columns of Table 7-2 provide informa- In an effort to clarify the depositional context and in- tion on the thickness of the deposit between the dates, tegrity of the cultural zones in Blocks 2 and 3 at and the accumulation rate. We use the midpoint of the 41MM340, a series of snail and mussel shells were 2 sigma range as a time estimate, and the elevation selected, in consultation with TxDOT, and submitted difference, in cm, between dates as a thickness esti- for amino acid racemization. Detailed analysis of mate. While a number of assumptions are involved, amino acid racemization values on snail and mussel especially with the Block 3 data where the radiocar- shell may provide an inexpensive measure pertinent bon dates come from different units within the block, to assessing integrity of deposits (e.g., Ellis et al. 1996; the accumulation rate provides a gross estimate of Ellis 1997; Goodfriend 1987; Quigg 1999). Specifi- deposition rates and, by extension, the probability that cally, the racemization of amino acids L-isoleucine to deposits potentially represent a palimpsest. That is, D-alloisoleucine (A/I) in snails and mussels appears high accumulation rates are less likely to be over- to be primarily a function of time. While the racem- printed by multiple occupations, and thereby compli- ization rates are complicated by the temperature his- cate the interpretive potential of the deposits. An tory of a sample, and while different genera of examination of the Block 2 data suggest that the high- mollusks probably have different rates of amino acid est accumulation rate occurs in the area represented racemization, radiocarbon dating on snails, primarily by Zones 4, 5, 6, and 7, while Zone 2 has the lowest Rabdotus, suggests that shells of similar age have simi- accumulation rate. Within Block 3, the highest accu- lar A/I values (see Ellis et al. 1996). Ellis et al. mulation rates are present in Zone 7, with Zone 3 hav- (1996:191) state that “the relative contemporaneity of ing the lowest accumulation rate. artifacts in a deposit will be approximated by the rela- tive contemporaneity of snails buried in the same de- The radiocarbon dates clearly suggest that the sedi- posit.” That is, the A/I values on shell can be used as a ments in both Blocks 2 and 3 accumulated over a rela- proxy for assessing assemblage integrity. When shells tively brief period of time. While a number of of radically different ages, as measured by different interpretations of these data are possible, it is likely A/I values, are present in the same deposit, either burial that the sediments accumulated in no more than 1,100 of the deposits occurred at a slow rate or the deposits

35 have been turbated after deposition. In their view, ei- mussel data are discussed subsequently, all ratios of ther situation would result in an assemblage of lim- aspartic acid to glutamic acid are relatively uniform ited integrity and, by extension, limited utility for and cluster around a value of 1.1. The lack of outliers addressing “behavioral” questions (Ellis et al. 1996). in these data suggests that the snails submitted for ra- cemization were not exposed to heat. Finally, note that Given previous work on Rabdotus racemization, A/I values for either Polygyra mooreana or samples of this genus would have been ideal. Unfor- Praticolella berlandieriana could not be located in tunately, Rabdotus was not represented from all pro- the literature. However, if A/I values from the same veniences in sufficient quantity for analysis (see genera are consistent, then establishing the exact rate Appendix G). Consequently, 25 samples of Polygyra of change is less important than examining the clus- mooreana and seven samples of Praticolella tering of A/I values for a given provenience. If values berlandieriana were selected from the four primary cluster, then it is likely that they date to the same time zones within the two blocks. In addition, 32 samples period. of mussel shell, consisting of 27 samples of Amblema plicata and five samples of Tritogonia verrucosa, were Figure 7-12 presents the snail A/I values in Blocks 2 also submitted. The samples represented both Block and 3. For comparative purposes, we include the A/I 2 and 3 at the site, with the majority (n=48) coming values published by Ellis et al. (1996:201) for from Block 3. TxDOT supplied CAR with the results Rabdotus from alluvial proveniences on Fort Hood, of the analyses. These are presented as Appendix I. located about 70 km to the northwest of the current

Assessing the degree to which a given provenience is disturbed ,5 through A/I values is minimally ]RQHÃ  dependent on having some idea ]RQHÃ  of the time represented by a ]RQHÃ  given A/I value, a variable that ]RQHÃ  is probably genera dependent,   some idea of the measurement V H  F  error in determining the A/I Q LH   value itself, and an understand- Q H ing of the temperature history of Y   R the snails submitted. Goodfriend U   OÃ3 (2000:405) reports that the ana- LD Y   lytical error is approximately OX O  Ã$  five percent. In addition, Ellis G R   (personal communication, 2000) R Ã+ suggests that ratios of two other W   ) amino acids, aspartic acid and  

glutamic acid, may provide a   measure of the impact of heat-   ing in both snails and mussels, as aspartic acid has a lower tem-      perature threshold for racemiza- $PLQRÃ$FLGÃ5DFHPL]DWLRQÃ $, Ã9DOXHV tion. Appendix I presents the ratio of aspartic acid to glutamic 5DEGRWXVÃ )RUWÃ+RRG 2QO\Ã3RO\J\UD 3RO\J\UDÃ 00 2QO\Ã3UDWLFROHOOD acid, as well as the associated A/ 3UDWLFROHOODÃ 00 I values, for both snails and mus- sels from the site. While the Figure 7-12. Amino acid racemization (A/I) values for Fort Hood proveniences (Ellis 1997) and 41MM340.

36 project area. Note that values greater than 0.35, which snails from roughly the same age are being distributed Ellis et al. (1996) suggest are indicative of heating, throughout the profile by turbation, or additional fac- have been eliminated from the comparisons. Values tors are impacting the A/I values. Each of these possi- in this range were not present in the 41MM340 snail bilities is considered below. data, though several of the Fort Hood proveniences had these high values. Examination of the 41MM340 As noted in Appendix G, the snail assemblage at the data in the figure will suggest that: 1) variability within site was in excellent condition. The relatively low a given provenience is small; 2) within each prove- number of species present and the occurrence of only nience there is a clustering of the A/I values from a one species of aquatic gastropod suggest that the snail given species; 3) the clustering of values in the assemblage was deposited in situ and was not the re- 41MM340 data are comparable to the clustering re- sult of fluvial transport. There is, however, a change flected in many of the proveniences from Ft. Hood; in the relative abundance of different species. and 4) there is no directional change from the Zone 8 Rabdotus dealbatus are in abundance in the deepest data through the Zone 2 data in 41MM340. cultural zones, while Helicina orbiculata dominate the middle and upper zones. Examination of the relative Within Zones 2, 4, and 6 at 41MM340, the spread of weights of snails by zone within each block (see Fig- A/I values on Polygyra mooreana, the species that ure 7-13) will demonstrate that these changes in spe- dominates samples from these zones, are relatively cies are reflected by changing weights. While a variety small. This is especially the case in Zones 2 and 4. of factors may account for the distribution of differ- Zone 6 has two clusters, one consisting of three val- ent species weights seen in the figure, the pattern is ues centered at 0.028 and a second three-value cluster not a homogenous distribution that might result from around an A/I value of about 0.06. The clustering in extensive post-depositional turbation. While some the upper zones at 41MM340 are consistent with clus- lower level of turbation may be consistent with the ters from several proveniences at Ft. Hood that, on patterns seen in Figure 7-13, it is unlikely that the lack the basis of other data, Ellis et al. (1996) suggested of directional patterning in A/I values can be accounted had high integrity. In Zone 8, the six samples of for solely by reference to extensive post-depositional Praticolella berlandieriana have a wider spread, though turbation. three of the samples cluster around a value of 0.058. Ellis (personal communication 2001) suggests that the To the far right of Figure 7-12 we provide the lack of a consistent patterning in the deposits may be interquartile range (IR) for the A/I ratios for all 16 related to temperature differences. As noted above, proveniences. The IR provides a comparable measure temperature can have an impact on the A/I ratios, with for the clustering, identifying the range in which 50 exposure to high temperatures resulting in higher rates percent of the cases are located. The IR values, which of racemization, and higher A/I values, and lower tem- range from a high of .1174 for provenience 9 at Ft. peratures resulting in lower A/I values. Thus, if tem- Hood to a low of .0120 for provenience 11 at Ft. Hood, peratures within a given level were uniformly reduced, reinforces the visual impressions that most of the A/I values would still cluster, but be out of place rela- 41MM340 cases are tightly clustered. These cases tive to other levels. Examination of Figure 7-12 sug- range from .0219 to .0290. gests that this is precisely the case with the Zone 6 samples, where there is reasonably good clustering of While values cluster within several of the zones for the the values, but a lower overall value relative to Zones 41MM340 data, taken as a whole there is no apparent 4 and 2. Possible reasons for the reduced temperature increase in A/I values with decreasing depth for the 23 pattern may include fluctuations in the water table or values of Polygyra mooreana from these upper zones. changes in surface hydrology resulting in saturation Nevertheless, based on the radiocarbon results, the bot- of the soil, either of which could result in lower over- tom of Zone 6 in Block 3 has a radiocarbon date with a all temperatures. However, we have no detailed infor- 2 sigma midpoint at about 1115 B.C. (1260–970 B.C.) mation on either of these potential causes operating at while the top of Zone 2 in Block 2 has a midpoint date 41MM340. of 570 B.C. (770-370 B.C.). It appears, then, that either

37 In addition to the snail shells, 32 samples of mussel As noted above, the ratio of aspartic acid to glutamic shell, consisting of 27 samples of Amblema plicata acid may provide a measure of the impact of heating and five samples of Tritogonia verrucosa, were also in snails and mussels. Reference to the mussel shell submitted from 41MM340 for analysis (see Appen- data in Appendix I will demonstrate that in several dix I). As with the snail data discussed above, the ra- cases, extremely high A/I ratios are associated with tios of L-isoleucine to D-alloisoleucine (A/I) were low values in the ratio of aspartic acid to glutamic monitored. The assumptions are identical to those for acid. Most of the lower aspartic acid to glutamic acid the snail data. However, previous work on A/I values cases have A/I values in excess of 0.2. Given Ellis’ of mussels has suggested that heating, noted as a po- (1997) previous results, we will eliminate those seven tential problem in snails, is common in these larger cases with A/I values greater than 0.2 from the analy- mollusks. In an analysis of Amblema plicata Ellis sis (see Appendix I). However, it may still be the case (1997) has demonstrated that a substantial number of that some unknown component of the remaining mus- samples from a burned rock midden have A/I values sel A/I values have been altered by exposure to heat, a that are in excess of 0.20, values that are likely the factor that would both inflate individual values and result of heating. change the overall distribution.

Figure 7-14 presents the A/I ratios for the remaining 25 samples. Note that, as with the snail data discussed earlier, in most cases the mussels are tightly clustered.

 Only a single value, in Zone 8, is an obvious outlier relative to the rest of the distribution. In addition, as  with the snail data, the A/I values in the mussels show  V no overall trend in increasing values with depths. H Q R  Ã= While a trend is apparent in the Block 3 data sets LF K S  consisting of Zones 8, 6, and 4, the Zone 2 data, that UD J WL comes from Block 2 at 41MM340, do not follow the D  WU 6 trend. As with the snail data, a variety of factors may 



         7RWDOÃ6QDLOÃ:HLJKWÃ JU 





  R  WL D Ã5   , $ V  H  Q R Ã= LF   K S UD   J WL D WU  6     

       =RQH 7RWDOÃ6QDLOÃ:HLJKWÃ JU $PEOHPDÃSOLFDWD 7ULWRJRQLDÃYHUUXFRVD Figure 7-13. Total snail weight (gr.) by zone for Block 2 (top) and Block 3 (bottom) at Figure 7-14. A/I ratios (less than 0.2) for mussel shells 41MM340. by zone at 41MM340. 38 account for these differences, but differential heating Texas form. The Lange point has a slightly expanding is the most likely candidate for both the extreme outlier stem and a convex base (Figure 7-15c). It is relatively value in Zone 8, as well as the ambiguity of the overall thick and roughly shaped. Lange points are typically pattern of A/I values on mussels. common in Central Texas. The Morrill point has a parallel to slightly barrel-shaped stem and a straight base (Figure 7-15d). It is generally thought of as an Temporal Diagnostic Chipped Stone Tools East Texas point type (Suhm and Jelks 1962:223). The Marshall point has a short, slightly expanding stem, A total of 14 projectile points were identified in the and a concave base (Figure 7-15e). Its short barbs are 41MM340 collection. In addition, four Erath bifaces slightly downward pointing. It is a point type com- recovered from excavation can be used to provide mon to Central Texas and the Edwards Plateau. The broad temporal information on specific archaeologi- two Castroville points have broad expanding stems cal components within the site (Table 7-3). and straight bases (Figure 7-15f). This type is also most frequent on the Edwards Plateau and in Central Texas. The Ellis point, the final typed specimen, has Of the fourteen dart points eight can be typed, two are an expanding stem, convex base, and short downward untyped, and the remaining four are untypeable frag- pointing barbs (Figure 7-15g). It is made of light gray ments. The eight typed specimens consist of two Gary, chert similar to that obtained in the vicinity of a Lange, a Morrill, a Marshall, two Castroville, and Georgetown, Texas. The type is common in both East an Ellis (Figure 7-15). One of the two Gary points and Central Texas. (Figure 7-15a) is made of brown, relatively poor qual- ity, petrified wood. The second Gary is made of natu- rally pinkish-red chert (Figure 7-15b) that is a common A complete dart point with a narrow parallel stem and component of upland gravel concentrations along the a straight base cannot be assigned to a specific type western edge of East Texas. The type is a typical East (Figure 7-15h). The longitudinally asymmetrical blade

Table 7-3. List of potential chronologically diagnostic artifacts recovered from 41MM340

Lot Block Unit Zone Level Depth Artifact Category Comments 42 1 A n/a 6 50-60 cmbs Morrill Dart Point Complete; Figure 7-15d Stem fragment; Bulverde-like but 97 1 A n/a 10 90-100 cmbs Untyped Parallel Stemmed Straight Based shorter than typical stem 122 1 C n/a 10 90-100 cmbs Lange Dart Point Complete; Figure 7-15c 87 2 D 2 1 @56 cmbs Castroville Dart Point Proximal fragment; Figure 7-15f 220 1 D n/a 12 110-120 cmbs Castroville Dart Point Stem fragment 224 2 C 7 3 @110 cmbs Marshall Dart Point Complete; Figure 7-15e Untyped Narrow Parallel Stemmed Straight Complete; may be reworked Montell; 294 2 B 8 2 140-150 cmbs Based Figure 7-15h Complete; made on “Georgetown flint” 66 3 A 2 2 50-60 cmbs Ellis Dart Point Figure 7-15g 89 3 D 3 3 60-70 cmbs Gary Dart Point Proximal fragment; Figure 7-15a 212 2 D 6 2 100-110 cmbs Gary Dart Point Proximal fragment; Figure 7-15b 152-1 3 A 4 2 84-94 cmbs Untypable Stem Fragment Small stem fragment 89-1 3 D 2 3 20-30 cmbs Untypeable Dart Point Distal Fragment Manufacture broken preform 157-1 1 B n/a 11 100-110 cmbs Untypeable Dart Point Distal Fragment Use broken 196-1 1 D n/a 11 100-110 cmbs Untypeable Dart Point Distal Fragment Use broken 111-1 2 B 4 1 @67 cmbs Large Erath Biface Complete; Figure 7-15i 220-2 1 D n/a 12 110-120 cmbs Erath Biface Preform Overshot distal fragment Distal fragment with evidence of 137-2 3 D 4 2 85-89 cmbs Erath Biface Fragment rejuvenation 137-1 3 D 4 2 @92 cmbs Small Erath Biface Complete 39 Figure 7-15. Dart points and Erath biface recovered from 41MM340. a, b) Gary; c) Lange; d) Morrill; e) Marshall; f) Castroville; g) Ellis; h) untyped parallel stemmed point; i) Erath biface. and the reworking along one edge of the stem and blade II (see Johnson 1995:Figure 80), is from 110 cm bs, in suggest that the point might be a reworked Montell Block 2. Regardless of their specific and relative strati- specimen. However, the stem is somewhat wider than graphic positions, based on their previously known the split stem of a typical Montell point. A short chronological affiliation, the six types are consistent untyped rectangular stem fragment is reminiscent of with the range of site occupations as reflected by the Bulverde points, although it is somewhat shorter than 14C dates (i.e., later half of the Late Archaic II through typical for the type. The remaining projectile point the early half of the Late Archaic I). fragments are untypeable distal (n=3), and proximal (n=1) fragments. The four Erath bifaces from the site can be assigned to the Late Archaic period (see Story and Shafer The relative stratigraphic position of the eight typed 1965:45). Two of the four specimens are complete. specimens is somewhat confusing with Gary points The larger of the two complete Erath bifaces (Figure ranging between 60–70 cm bs in Block 3 and 100-110 7-15i) has a morphology not unlike that of the Clear cm bs in Block 2, and the two Castroville points rang- Fork gouges (Story and Shafer 1965; Prewitt 1982:Fig- ing from 56 cm bs in Block 2 to 110–120 cm bs in ure 36). Were it not for the two notches on the lateral Block 1. On the other hand, the Ellis point that tends edges of the specimen near the working edge, it would to be assigned to the later part of Late Archaic II is at have been classified as a Clear Fork tool. The smaller 50–60 cm bs in Block 3, and the Marshall, that tends complete specimen has a similar morphology with to precede the smaller dart points of the Late Archaic deep notches in the vicinity of the working edge and

40 one concave lateral margin. The four Erath bifaces ground surface, approximately 15 cm (6 in.) above are distributed between 67–120 cm bs, with three of the underlying rich cultural zone. The slab is encased the four coming from between 67–92 cm bs. with remnants of the deteriorated exterior surface ex- foliated into the surrounding soil, forming a tan, sandy The small collection of formal tools exhibits a mix of sediment. Since it is contained within the upper zone artifacts typically thought to be more closely affili- of relatively recent fill and appears to be unmodified, ated with Central Texas (e.g., Castroville, Marshall, it is probably non-cultural in origin. Ellis) or East Texas (e.g., Gary, Erath bifaces). Inter- estingly, with the exception of the Ellis point, arti- Feature 2 facts that date roughly to the latter part of the Late Feature 2 is a rock-lined basin hearth and consists of Archaic I seem to show a stronger affiliation with East a roughly circular concentration of fire-cracked rock Texas than artifacts that date to the early part of the with associated fired clay, charcoal, freshwater mus- Late Archaic I and later portion of Late Archaic II. sel shell, and lithic debitage. The feature was encoun- This pattern may signal some interesting regional tered within Zone 4 of Block 3, XUC and XUD at changes in mobility and resource procurement. The approximately 83 cm (33 in.) below ground surface chronological divisions referred to here accord with (Figure 7-16). The cobbles comprising the exterior por- Johnson and Goode (1994). tion of the feature averaged approximately 10 cm (4 in.) in diameter. The concentration measures approxi- mately 85 cm (34 in.) in diameter and appeared to Feature Descriptions exhibit a slight vertical concavity.

A total of seven cultural features were identified and Feature 3 excavated at 41MM340. Six features were newly re- Feature 3, a refuse pit, was encountered during the re- corded and one (Feature 6) was previously recorded excavation of backhoe trench (BHT) 26. The feature during the survey phase. Five of the encountered fea- occurs at approximately 110 cm bs within the upper tures (Nos. 2, 4, 5, 6, and 7) can be classified as true, aspect of a weakly developed paleosol. The pit is coni- rock-lined basin hearths with associated fire-cracked cal in shape, with the upper rim measuring approxi- rock, charcoal, and other heat-altered remnants (Table mately 60 cm in diameter and the base approximately 7-4). A single feature (No. 3) is interpreted as a refuse 20 cm in diameter. The overall depth of the pit is 25 pit. More detailed descriptions of the individual fea- cm. The rich, midden-stained soil of the refuse pit in- tures are discussed below. cluded burned bone, chipped stone, mussel shell, and small fragments of burned rock. The refuse pit, Fea- Feature 1 ture 6 (see below), and a nutting stone (catalog num- Feature 1 is a relatively large (>50 cm) slab of friable ber 157-2) were all encountered within a less than 70 sandstone with no associated artifacts. This anomaly cm radius contained in the upper aspect of the paleosol. was encountered in the northwestern quadrant of Block Excavation of the three arbitrary levels in which these 3, Excavation Unit (XU) A at 40 cm (16 in.) below features occurred resulted in the recovery of roughly

Table 7-4. Features encountered during manual excavation at 41MM340

Feature Block Unit Elevation Type 1 3 A 33-42 cmbs Large, unmodified sandstone slab. Probable non-feature. 2 3 C,D 83-89 cmbs Rock-lined hearth. 3 1 B 100-135 cmbs Unlined pit feature. 4 2 C 78-86 cmbs Rock-lined hearth. 5 3 A 85-94 cmbs Rock-lined hearth. 6 1 A 113-125 cmbs Rock-lined hearth. 7 3 A,B,C,D 92-111 cmbs Rock-lined hearth.

41 one third of the artifact assemblage for the entire block, indicating an intense use/activity area.

Feature 4 Feature 4 is a rock-lined basin hearth and consists of an irregular, oblong concentration of fire-cracked rock with associated charcoal, sparse freshwater mussel shell, and moderate to abundant lithic debitage. The feature was encountered within Zone 4 of Block 2, XUC at approximately 78 cm (31 in.) below ground surface. A majority of the cobbles contained within the hearth were tabular burned sandstone 5–10 cm in diameter. Also present within the feature were several siliceous pebbles, averaging approximately 3 cm in diameter (Figure 7-17). Some of these pebbles ap- peared to have been exposed to heat, suggesting they may have been utilized as boiling stones. The concen- tration measures approximately 40 cm (16 in.) wide and 50 cm (20 in.) long and appeared to exhibit a slight vertical concavity.

Feature 5 Figure 7-16. Feature 2, Block 3, 41MM340. Feature 5 is a large rock-lined basin hearth and con- sists of an ovate concentration of fire-cracked rock with associated charcoal. The feature was encountered within Zone 4 of Block 3, XUA at approximately 85 cm (33 in.) below ground surface. A small portion (less than 10 percent) of the feature appears to extend north and west of XUA and was not excavated. The con- centration measures approximately 55 cm (22 in.) at the widest point and 81 cm (32 in.) at the longest point and appeared to exhibit a slight vertical concavity.

Feature 6 Feature 6 was originally recorded during the TxDOT survey phase in BHT 26 as a burned rock feature (Ahr and Abbott 1999:11). Re-excavation of the backhoe trench allowed relocation of Feature 6 as well as the associated Feature 3 (see above). Feature 6 is a small rock-lined basin hearth consisting of a circular con- centration of fire-cracked rock. Abundant charcoal was encountered in the matrix immediately beneath the concentration of burned rocks. The feature was en- countered in Block 1, XUA at approximately 113 cm (44 in.) below ground surface. The extant portion of the feature suggests an original diameter of approxi- Figure 7-17. Feature 4, Block 2, Unit C, 41MM340. mately 40 cm (16 in.) and a depth of 12 cm (5 in.).

42 Feature 7 selected artifact and ecofact classes for the zones en- Feature 7 is a rock-lined basin hearth and consists of countered, given the vertical undulations and variabil- an irregular concentration of fire-cracked rock with ity of zone thickness. The data in Tables 7-8 and 7-9 associated charcoal, and sparse freshwater mussel are adjusted for a volume of 100 cubic cm. Many of shell. The feature was encountered within Zone 6 of these distributions have been presented previously (see Block 3, XUA, XUB, XUC, and XUD at approxi- Figures 7-5 through 7-8, Figure 7-13). An examina- mately 92 cm (36 in.) below ground surface. The tion of the other categories in Table 7-5 for Block 1 cobbles comprising the exterior portion of the feature will reinforce the pattern of at least two distinct modes averaged approximately 7 cm (3 in.) in diameter, with suggested by the chipped stone distribution presented smaller cobbles and pebbles located in the interior in Figure 7-5. In addition, the distribution of all classes portion of the hearth. The concentration measures ap- of material in Blocks 2 and 3 (see Tables 7-8, 7-9) proximately 65 cm (26 in.) wide and 75 cm (30 in.) support the presence of multiple zones in each of the long and exhibited vertical concavity of approximately two blocks. 19 cm (7 in.).

Debitage Analysis Artifact Distribution All unmodified debitage recovered from the three Tables 7-5 through 7-7 present the distribution of ar- blocks at 41MM340 has been categorized in terms of tifacts and ecofacts recovered from the three blocks corticate (i.e., primary and secondary) and decorticate excavated at 41MM340. In addition to the raw data (tertiary) specimens. A comparison of relative propor- counts for each category, volumetric calculations are tions of these two categories can identify differences provided for the eight zones in Block 2 (Table 7-8) in stages of reduction either between or within blocks and the nine zones in Block 3 (Table 7-9). These volu- and cultural zones. metric tables provide a standardized comparison of

Table 7-5. Vertical Distribution of Artifacts and Ecofacts Recovered from Block 1, 41MM340

Chipped Mussel Shell Mussel Shell Snail Shell Level Stone FCR Umbo Fragments Snail Shell Fragments Bone N N Wt (gr.) N Wt (gr.) Wt (gr.) N Wt (gr.) Wt (gr.) N Wt (gr.) 01 2 3 22.77 2 6.5 0.69 0 0 0 0 0 02 2 3 0.73 0 0 0 2 0.3 0 62 28.82 03 6 5 112.4 1 1 1 46 9 0 2 0.34 04 7 0 0 0 0 0 23 5 0 0 0 05 16 9 77.94 5 13.6 223.4 350 73.3 1 6 2.32 06 83 37 430.21 27 35.4 11.39 588 104 189.98 11 12.21 07 417 322 1289.92 89 209.7 33.53 356 61.8 15.65 55 15.22 08 159 101 1492.65 26 40.1 8.12 477 98.1 18.4 46 6.44 09 75 35 401.12 39 92.9 612.35 569 101.2 1.65 6 1.97 10 418 160 891.62 158 427.8 96.08 1302 225.1 9.66 29 12.91 11 300 183 1479.63 57 209 25.82 569 104.72 5.86 4 0.56 12 405 90 218.67 50 104.3 44.45 362 51.9 5.87 8 0.74 13 60 7 23.89 7 29.7 5.64 200 36.3 3.66 2 0.19 14 12 1 0.3 11 27.6 3.76 276 45.1 1059.95 3 1.23 15 6 34 193.24 1 107 4.55 186 34.4 2.9 1 0.08 16 7 3 3 0 0 3.13 97 18.7 1.17 0 0 Total 1975 993 6638.09 473 1304.6 1073.91 5403 968.92 1315.75 235 83.03 43 Table 7-6. Vertical Distribution of Artifacts and Ecofacts Recovered from Block 2, 41MM340

Chipped Mussel Shell Snail Shell Zone Stone FCR Mussel Shell Umbo Fragments Snail Shell Fragments Bone 1 105 158 809.8 151 747.5 134.22 1796 304 3.88 12 0.17 2 432 779 4238.73 1414 16365.14 244.46 725 151.09 4.74 89 19.09 3 49 91 606.49 8 31.7 8.14 35 18.7 4.9 7 2.27 4 1636 5348 37305.29 313 1211.7 416.79 1607 332.4 440.4 224 103.67 5 254 325 2667.13 25 87.3 67.91 409 73.3 10.6 38 14.29 6 459 481 2019.08 70 148.4 64.03 558 105.7 10.23 21 3.95 7 453 238 738.64 136 350 160.87 814 155.7 179.45 31 6.98 8 1511 832 7492.8 655 1810.25 401.95 1927 1023.9 409.36 195 74.68 Total 4899 8252 55877.96 2772 20751.99 1498.37 7871 2164.79 1063.56 617 225.1

Based on the distribution of the number of chipped The breakdown of corticate and decorticate debitage stone artifacts by excavation level, two peaks, possi- within the eight zones of Block 2 does indicate some bly representing depositional zones, can be identified variability. The proportion of decorticate debitage in Block 1. The upper peak includes Levels 6–8, while ranges from as low as 70–71 percent in the two upper the lower peak consists of Levels 10–12. The tabula- zones (1 and 2), to the mid-seventies (76, 77, 77, and tion of cortex categories within these two samples of 75) in Zones 3, 4, 6, and 8, respectively, and as high debitage (n=644 upper zone; n=1107 lower zone) as 86 and 80 percent in Zones 5 and 7, respectively. shows no differences between them. Seventy percent The comparison of the numbers of corticate and deco- of the debitage is decorticate while 30 percent is cor- rticate debitage within these zones using adjusted stan- ticate in both samples. These percentages are roughly dardized residuals indicates that tertiary debitage is representative of full reduction sequences that progress under-represented in Zone 2 and tends to be under- from locally available corticate nodules to finished represented in Zones 1 and 8, and is over-represented bifacial artifacts such as large projectile points and/or in Zone 5 and tends to be over-represented in Zone 7 bifacial knives. Although it is likely that finished tools (Table 7-10). were also rejuvenated on site, the debitage collection is not dominated by the products of the early or late The percentage of decorticate debitage within the nine stages of reduction. zones identified in Block 3 shows a range of 67–81 percent. Zones 1, 4, 8, and 9 have relatively low

Table 7-7. Vertical Distribution of Artifacts and Ecofacts Recovered from Block 3, 41MM340

Chipped Mussel Shell Mussel Shell Snail Shell Zone Stone FCR Umbo Fragments Snail Shell Fragments Bone N N Wt (gr.) N Wt (gr.) Wt (gr.) N Wt (gr.) Wt (gr.) N Wt (gr.) 1 277 119 634.29 128 616 84.18 691 126.3 4.54 48 28.83 2 211 420 2230.48 118 394 275 226 35 3 199 604.33 3 238 337 1199.57 66 137 81 383 89 2.25 26 6.79 4 806 3027 17786 976 2374 560 1193 265 105.28 172 49.64 5 114 56 857 214 2380 308 218 10 9 36 6 6 371 940 10535 388 1445 670 327 42 68 92 40 7 143 1276 7589 434 1301 237 856 444 102 28 6.81 8 444 582 5634 1154 4808 259 5297 2955.2 514 144 66.61 9 67 115 1099 54 206 36 327 136.72 26 3 0.3 Total 2671 6872 47564.34 3532 13661 2510.18 9518 4103.22 834.07 748 809.31 44 Table 7-8. Standardized Artifact and Ecofact Distribution, Block 2, 41MM340

Zone Chipped Stone FCR wt (gr.) Mussel wt (gr.) Snail wt (gr.) Bone Count 1 8.20 63.26 68.88 24.05 0.94 2 261.82 2568.93 10066.43 94.44 53.94 3 33.33 412.58 27.10 16.05 4.76 4 180.97 4126.69 180.14 85.49 24.78 5 62.56 656.93 38.23 20.67 9.36 6 126.80 557.76 58.68 32.02 5.80 7 39.05 63.68 44.04 28.89 2.68 8 141.21 700.26 206.75 133.95 18.22 Total 853.95 9150.08 10690.26 435.57 120.48 percentages of tertiary flakes. Zones 2, 3, and 5 have debitage in the other zones of the two blocks appear to the highest percentages of decorticate debitage in the be contradictory (see Zones 2 and 8). block. The comparison of the numbers of corticate and decorticate debitage within these zones, using In general, the reduction of nodules to finished bifacial adjusted standardized residuals, indicates that tertiary artifacts (e.g., large dart points and/or bifacial knives) debitage is under-represented in Zone 1 and 8, and tends to generate around 70 percent decorticate tends to be under-represented relative to corticate debitage. The higher percentages in Zones 3, 5, 6, and debitage in Zone 4 (Table 7-11). Tertiary debitage is 7 across the two blocks suggests that the debitage over-represented in Zones 2, and 3, and tends to be samples have a somewhat higher proportion of late re- over-represented in Zones 5 and 6. duction stage debitage than would normally be ex- pected. Such a pattern may derive from the utilization The comparison of these trends in the two blocks reveals of tertiary flake cores for tool manufacture and/or the that tertiary debitage tends to be under-represented in repair and rejuvenation of finished tools. The between Zone 1 and 8, and over-represented in Zone 5. In block differences in Zones 2 and 8 may reflect intra- addition, the percentages of tertiary debitage tend to be site structural differences in activity area use. high in Zone 3 (76 percent in Block 2, 81 percent in Block 3), Zone 5 (86 percent in Block 2; 79 percent in To investigate the possibility that biface and uniface Block 3), Zone 6 (77 percent in Block 2, 77 percent in manufacture may have been carried out in different Block 3), and Zone 7 (80 percent in Block 2; 77 percent proportions in the different zones of 41MM340, all of in Block 3). The relative percentages of decorticate the move than 2,600 pieces of unmodified debitage

Table 7-9. Standardized Artifact and Ecofact Distribution, Block 3, 41MM340

Zone Chipped Stone FCR wt (gr.) Mussel wt (gr.) Snail wt (gr.) Bone count 1 22.89 52.42 57.86 10.81 3.96 2 60.28 637.28 191.14 10.85 56.85 3 29.02 146.28 26.58 11.12 3.17 4 244.24 5389.69 889.09 112.2 52.11 5 54.28 408.09 1280 9.04 17.14 6 132.5 3762.5 755.35 39.28 32.85 7 30.42 1614.68 327.23 116.17 5.95 8 83.78 1063.01 956.03 654.56 27.16 9 5.58 91.58 17.19 13.56 0.25 Total 662.99 13165.53 4500.47 977.59 199.44

45 from Block 3 was categorized into distinct flake types Table 7-10. Comparison of debitage cortex categories (i.e., biface manufacture, biface thinning, uniface using adjusted standardized residuals, Block 2, manufacture and rejuvenation, core reduction, blade 41MM340 production, core/platform preparation, and notching flakes). The large proportion of the identified flake Zone Corticate Decorticate types (n=1,214, 46 percent) are core/platform prepa- 1 1.52 -1.52 ration flakes. Only seven (less than one percent) blades 2 2.76 -2.76 were identified in the sample and not surprisingly, only 3 0.2 -0.2 four (less than one percent) debitage were identified 4 -0.72 0.72 as clearly representing core reduction rather than 5 -3.6 3.6 biface manufacture. Six of the seven blades are from 6 -0.28 0.28 Zone 6. Thirty-two (1.2 percent) notching flakes were 7 -1.74 1.74 identified, with 15 (0.6 percent) occurring in Zone 6 8 1.52 -1.52 Bold numbers represent statistically significant adjusted or below (80 percent from Zone 6). Biface manufac- standardized residuals significant at the 0.05 level of sig- ture flakes constitute a total of 713 (27 percent) of the nificance (+/- 1.96; Everitt 1977:47). total sample, about half of these (n=355) are biface thinning flakes. Uniface manufacture flakes make up a small proportion of the sample (n=106, 4 percent). Table 7-11. Comparison of debitage cortex categories The comparison of the frequencies of biface manu- using adjusted standardized residuals, Block 3, facture, biface thinning, and uniface manufacture 41MM340 debitage by zone indicates that biface thinning flakes are under-represented in Zone 1 and over-represented Zone Corticate Decorticate in Zone 3 (Table 7-12). These flake types also tend to 1 2.46 -2.46 be under-represented in Zone 7, although not to a sta- 2 -1.97 1.97 tistically significant degree. Biface manufacture flakes 3 -2.89 2.89 (i.e., early and middle reduction stage flakes) have 41.83-1.83 the opposite tendency. Thinning flakes are under-rep- 5 -1.51 1.51 resented in Zone 3, and over-represented in Zone 7. 6 -1.74 1.74 There is also a tendency, although not statistically sig- 7 -0.96 0.96 nificant, for thinning flakes to be over-represented in 8 2.51 -2.51 Zone 1 and under-represented in Zone 8. Uniface 90.25-0.25 manufacture/rejuvenation debitage tends to be under- Bold numbers represent statistically significant adjusted represented in Zones 2, 4, 5, and 7, and over-repre- standardized residuals significant at the 0.05 level of sig- nificance (+/- 1.96; Everitt 1977:47). sented in Zones 6, 8, and 9.

The tendency for early (manufacture) and late (thin- ning) reduction stage debitage to exhibit contrasting The tendency for uniface manufacture/rejuvenation distribution patterns in Zones 1, 3, and 7 may reflect flakes to be over-represented in the deeper zones may differences in biface manufacture strategies. The over- suggest that activities necessitating unifacial tools such representation of biface manufacture (early and middle as scrapers (i.e., hide processing) and/or unifacial stage) debitage in Zones 1 and 7 may reflect the stage adzes (i.e., wood working) may have been more preva- manufacture of unfinished bifaces for transport to lent during the earlier occupations of the site. Although other localities. On the other hand, the over-represen- very preliminary, it may be worth further exploration tation of thinning flakes in Zone 3 may indicate an to determine whether these patterns signal a connec- emphasis on the manufacture of finished bifaces, po- tion of the inhabitants of the site to Blackland Prairie tentially indicating “gearing-up” for activities that are bison hunting adaptations. expected to necessitate large numbers of finished bifaces (e.g., projectile points and/or knives).

46 Table 7-12. Comparison of the occurrence of three Freshwater Mussel Shell flake types by zone using adjusted standardized residuals, Block 3, 41MM340 Robert G. Howells conducted analyses of freshwater mussel specimens recovered from three of the four Biface Biface Uniface cultural zones encountered at 41MM340 (Appendix Zone manufacture thinning manuf./rejuv. D). Samples of 50 specimens selected to contain a 11.87-2.54 0.99 variety of species from Zones 2, 6, and 8 were sub- 2 0.27 0.86 -1.66 mitted. According to his report, all specimens were in 3 -2.29 2.35 -0.09 very good condition, with seven species identified, 4 0.64 0.47 -1.64 representing the Family Unionidae. Threeridge 5 0.43 0.82 -1.8 (Amblema plicata), the most abundant unionid taxon 6 -1.33 0.01 1.95 within the Central Brazos River Drainage, displayed 7 2.12 -1.34 -1.15 a marked increase from Zone 8 (124–134 cm bs) to 8 -0.41 -0.88 1.9 Zone 2 (43–53 cm bs). Conversely, smooth pimpleback 9 -0.88 0.01 1.27 Bold numbers represent statistically significant adjusted (Quadrula houstonensis) exhibited a proportionately standardized residuals significant at the 0.05 level of sig- greater decrease between the two zones. As freshwa- nificance (+/- 1.96; Everitt 1977:47). ter mussels are sensitive environmental barometers, one possibility for this change in mussel population could be the regional climatic change toward a more arid environment. Despite this population variation, all the taxa represented in the archaeological assem- Vertebrate Faunal Material blage evolved in flowing waters and are most often associated with gravel, stable sand, and mud substrates Barbara A. Meissner, Center for Archaeological Re- in permanent-water conditions. No species represent- search, The University of Texas at San Antonio pro- ing slow moving, soft-bottomed stream conditions vided analyses of vertebrate faunal material recovered were identified. These data suggest the site was for- from the three excavation blocks at 41MM340. Ac- merly in relative close proximity to a perennial stream, cording to her report (Appendix C), the predominantly possibly a former meander of the Little River, and oc- fragmented bone is generally in moderate to good con- cupation may have ended before the meander was dition, with five genera of mammals, one genus of entirely cut off. reptile, and one genus of fish identified. The majority of the unidentified bone appeared to be from animals of the deer/antelope size category; however, based on Macrobotanical cortical thickness, at least two unidentified bones could have come from a bison-sized animal. Additionally, highly fragmented, unidentifiable specimens of bird, J. Philip Dering, Associate Director for Macrofossil turtle, and fish were present. There is little evidence Analyses, Paleoethnobotany Laboratory, Texas A&M of extensive atmospheric or chemical weathering, sug- University performed macrobotanical analyses of flo- gesting clayey, fluvial deposits buried the material tation samples recovered from each of the four cul- shortly after initial fragmentation and subsequent tural zones encountered within Block 3 (Appendix E). deposition. Overall, a relatively small amount of fau- According to his report, the archaeobotanical assem- nal material was encountered within these otherwise blage, both in quantity and size of the material, is some- rich, cultural deposits. In addition, a majority of the what reduced. The four samples submitted for analysis bone encountered is highly fragmented, suggesting the yielded only 16 very small (1–3 mm) wood fragments material was heavily processed. Further analyses may (Quercus sp. and Ilex sp.). Although the provide a clearer interpretation of the relative paucity archaeobotanical assemblage is somewhat reduced, of remains and the extent to which the remains present that botanical data have yielded some promising re- were processed. sults. Nut fragments (pericarp fragments) were recov- ered from three of the five (60 percent) flotation

47 samples examined. At least two nut types, thin (cf. open to sparsely wooded habitats again dominate the pecan) and thick (cf. hickory), were identified in the uppermost cultural zone. Overall, there appears to be assemblage. Due to the overall paucity of botanical a trend toward a drier environment through time as remains exhibited in the general level samples sub- represented within the four cultural zones. Alterna- mitted, future analyses should be directed toward tively, the shifts in dominant species through time samples recovered from heating features, which are might reflect changes in vegetation cover on the flood- conducive to the preservation of botanical materials, plain of Little River. The gastropod shells examined and from samples recovered for 14C dating since they appear to be deposited in situ and not from fluvial contained the bulk of the large pieces of carbonized transport. The relatively low number of species present plant remains. and the occurrence of only one species of aquatic gas- tropod support this. Furthermore, there appears to be little difference in the frequency of specimens encoun- Pollen and Phytolith tered among the cultural zones. These data suggest that each cultural zone represents a relatively short Analyses of pollen and phytolith samples extracted time span with a low snail population density. from each cultural zone were performed by John G. Jones, Associate Director for Pollen Analyses, Palynol- ogy Laboratory, Texas A&M University (Appendix Diatom F). Pollen and phytolith specimens were noted within the samples submitted, however, the overall state of Barbara Winsborough, of Winsborough Consulting, preservation was poor. The pollen concentration val- conducted analyses of four sediment samples and four ues within the four cultural zones are, from upper to fired clay samples from site 41MM340 (Appendix H). lower strata: 2,250 grains/ml, 936 grains/ml, 1,017 In addition, three modern comparative samples –wa- grains/ml, and 1,631 grains/ml. Environmental fac- ter from Little River, bottom sludge from Little River, tors such as cyclic wetting and drying in combination and topsoil from modern slough– were also examined with the acid soils (8.4 pH) of the region are not con- for diatom assemblages. The sediment samples were ducive to the preservation of pollen or phytoliths, and, collected from each of the four cultural zones in Block as such, the overall yield, or concentration value of 3 and BHT A. Evidence of diatoms was absent in the the individual samples do not lend themselves to reli- archaeological samples. The other half of the samples able interpretations. consisted of fired clay nodules recovered from vari- ous hearth features throughout the vertical column. Three of the four clay nodule samples examined con- Land Snail Shell tained diatoms. Positive samples recovered from lower strata contained only a few aerophilic diatoms (found Richard W. Fullington conducted analyses of gastro- in soil and mud), indicating that, while the fired clay pod samples extracted from each cultural zone (Ap- nodules do preserve diatoms to an extent, the nature pendix G). The samples consisted of snails recovered of the flood deposits normally preclude preservation from light and heavy fraction flotation and from ¼- of diatoms. However, a single sample associated with inch screen. According to his report, all specimens Zone 2, recovered in association with a hearth feature were in excellent condition, with thirteen species iden- extending just beneath the zone in Block 3, was the tified, representing ten families. The species repre- most diatomaceous, with 500 cells counted. The sented throughout the vertical column suggest little sample contained 25 species, representing a mixture variance in the environmental conditions within the of aerophilic diatoms and benthic aquatic species immediate area. Nonetheless, snail species preferring found attached to emergent vegetation and macroalgae. open and/or sparsely wooded habitats (e.g., Rabdotus The presence of this assemblage of diatoms suggests dealbatus dealbatus) are in abundance in the two deep- that they may have been transported to the site from a est cultural zones, woodland habitat preference spe- nearby stream, possibly attached to mussels or other cies dominate the middle zones, and species preferring edible material.

48 Summary Cultural Stratigraphy

Our limited hand testing at 41MM340, coupled with Similar to the northern section of 41MM340, this site extensive augering and supported by backhoe trench- initially exhibited no obvious cultural stratigraphy to ing has demonstrated that the site potentially contains warrant excavation using cultural strata. As such, a variety of data sets with relevance to a small portion manual excavation of the two blocks proceeded in ar- of the Late Archaic. In Block 1, while distinctive strata bitrary 10-cm (4-in.) levels throughout the 1.6-m (5.25- could not be identified, a consideration of artifact dis- ft.) vertical column. tribution clearly suggests that two, and perhaps three, distinct zones may be present. A series of radiocarbon During post-field analysis, based on the changes in dates on the feature recorded in BHT 26 suggests that artifact densities by level, the distribution of diagnos- some of this occupation occurred at around 970 B.C. tic chipped lithic artifacts, feature concentrations, and This is within the time frame reflected by radiocar- radiocarbon assays, it was determined that three cul- bon dates for both Blocks 2, and 3, where a series of tural zones could be discerned within the two blocks occupations that appear to date from about 1400 to of the site. Zone 1 is represented by the upper 50–70 450 B.C. were discovered. Not only do all three blocks cm of deposits across Block 1. A majority of the up- have data that are from a small span of the Late Ar- per aspect of Zone 1, that portion containing the Toyah chaic, but in both Blocks 2 and 3, a series of cultural Phase, was likely removed with mechanical excava- strata could be defined which subdivide this roughly tion of ¾ of the excavation block. The zone is charac- 1,000 year accumulation into units that, in many cases, terized by low artifact densities, few chipped lithics, are within the standard deviation of the associated a sparse scattering of mussel shell, and thirteen ab- radiocarbon dates. While it is currently impossible to original ceramics. A small circular unlined pit feature assign precise time limits to the various deposits within (Feature 1) was also within this zone. In Block 2, Zone the Block 2 and 3 testing, the various cultural layers 1 is manifested by a low density of chipped lithics appear to have accumulated rapidly, and certainly con- overlying Zone 2. Zone 2 in Block 1 consisted of a tain a variety of data sets relevant to questions of both 15–20 cm thick discrete shell lens containing large adaptation and change during the Late Archaic. numbers of burned rock, mussel shell, and chipped lithics. A lithic reduction station (Feature 3) and a rock- lined hearth (Feature 5) were also part of the cultural 41MM341 zone. In Block 2, this zone lacked the midden-like character. Although low frequencies of chipped lithics and mussel shell were present in the zone, hearth fea- Auger Data tures (Features 2, and 4) rather than refuse accumula- tions, characterized the zone. Burned rocks are Excavations of auger borings at 41MM341 revealed a abundant in the zone. Zone 3 in Block 1 begins be- single areal concentration of cultural material extend- tween 85–90 cm bs and underlies the material-rich ing the general width and approximately 50 m (164 ft.) midden zone. Low numbers of unmodified debitage, along the present and proposed rights-of-way (see Fig- mussel shells, and burned rock continue to a depth of ure 6-2). A total of 121 auger borings were excavated, 160 cm bs. In Block 2, this zone is characterized by encountering no evidence of former stream meanders moderate to high numbers of unmodified debitage, but in the form of wet soils or sediments. While a slough low numbers of burned rocks and mussel shells. exists to the south of the site bounds, there was no indi- cation in the auger boring excavations that this current slough has migrated through the site limits.

49 Table 7-13. Beta Analytic, Inc., results of radiocarbon dating for samples recovered from 41MM341

Aspect of Measured Corrected 13 12 Feature/ Radiocarbon C / C Radiocarbon Calibrated AD Calibrated AD Block Elevation Stratum Beta # Age Ratio Age Range at 1 Sigma Range at 2 Sigma 1 40-50 cmbs lower 141841 520 +/- 40BP -26.4% 500 +/- 40BP 1407-1440 1320-1480 1 70-80 cmbs lower 141842 1280 +/- 40BP -25.7% 1270 +/- 40BP 685-780 660-880 2 60-70 cmbs lower 141843 940 +/- 40BP -24.9% 940 +/- 40BP 1020-1160 1010-1190

Radiocarbon The single sample submitted from Block 2 was recov- ered from the lower aspect of a large (~200 cm diam- Beta Analytic, Inc., of Miami, Florida (Beta), pro- eter) hearth feature (Feature 4) in Zone 2. The hearth vided radiocarbon assays for three charcoal samples was encircled with stone and contained moderate recovered from 41MM341 (Figures 7-18 and 7-19; amounts of fired clay and fire-cracked rock and an Table 7-13). These were calibrated using the OxCal abundant amount of charcoal. The charcoal sample program (Ramsey 2000). Radiometric sampling (Beta produced a corrected date of 940±40 BP, which, when #141841) from Zone 1 in Block 1 produced a cor- calibrated, has a 1 sigma range of A.D. 1020–1160 (Beta rected date of 500±40 BP, which when calibrated #141843). Along with the single Scallorn arrow point yields a 1 sigma range of A.D. 1407–1440. This pe- recovered from the same level of the block, this date riod corresponds with the Perdiz Phase. The second places the feature within the Austin Phase of the Late sample submitted from Block 1 was recovered from Prehistoric Period. While the probable associated fea- 70–80 cm bs from the lower-middle portion of Zone ture (Feature 2) originally recorded in the survey-phase 2, the discrete shell lens. The sample (Beta #141842) backhoe trench (BHT 15) was not dated, charcoal re- dates the lower-middle portion of the discrete shell covered from BHT 17, located ~20 m to the north and lens at 1270±40 BP, with a calibrated 1 sigma range within the site limits, produced a corrected date of of A.D. 685–780. This date places the discrete shell 1050±50 BP (A.D. ~900) (Beta #128740), again consis- lens at the beginning of the Austin Phase. tent with the Austin Phase.

Figure 7-18. Profile of south wall of BHT 24 showing occupation zones and dates of radiocarbon assays, 41MM341.

50 sufficient amount of the stem remains to classify it as a Scallorn point (Figure 7-20a), an index marker of the Austin Phase. A Friday knife was recovered from the lower portion of Zone 2 (86 cm bs) in the same block. The well-flaked triangular biface has a slightly concave base and beveled and systematically serrated blade edges (Figure 7-20b). Friday knives tend to oc- cur in deposits spanning the Late Archaic II and the Austin Phase of the Late Prehistoric Period (Turner and Hester 1993). The third temporally diagnostic specimen is a manufacture failed roughly parallel- sided Hare biface (Figure 7-20c). It was recovered in an auger hole at a depth of about 20 cm bs. Hare bifaces may represent knife and/or dart point preforms and tend to occur in Late Archaic I assemblages (Prewitt 1982:Figure 37; Turner and Hester 1993).

Eleven of the 13 prehistoric ceramics were recovered at about 30 cm bs in Block 1. Based on construction method and temper, these specimens are likely to date Figure 7-19. Profile of east wall of Unit D, Block 2 to the later part of the Late Prehistoric Period (i.e., showing occupation zone and date of radiocarbon Toyah Phase). assay, 41MM341. The presence of ceramics at a depth of 30 cm bs, slightly above a radiocarbon assay of A.D. 1320–1480 Temporal Diagnostic Chipped Stone Tools (2 sigma range) from 40–50 cm bs in Block 1, is con- and Ceramics sistent with the Toyah Phase temporal assignment of Zone 1. The co-occurrence of the Scallorn arrow point Three temporally diagnostic tools were recovered from and Friday knife in the dense occupation zone dated 41MM341. In addition, 13 prehistoric ceramic frag- between A.D. 660–880 (2 sigma range) is consistent ments also were recovered from Block 1. Table 7-14 with the Austin Phase affiliation of the Zone 2 com- shows the distribution of the tools and ceramics by ponent. The possible Late Archaic II specimen (the block, unit, excavation zone, and level. The single ar- Hare biface) at a depth of 0–20 cm bs may be indica- row point is a medial fragment recovered from the tive of some degree of disturbance. Nonetheless, its middle portion of Zone 2 (60–70 cm bs) of Block 2. A

Table 7-14. Distribution of chipped stone tools and ceramics by block, unit, zone, and level, at 41MM341

Lot Block Unit Zone Level Depth Artifact Category Comments

55 2 C 2 7 60-70 cmbs Scallorn arrow point Medial fragment; Figure 7-20a 90 2 A 2 9 at 86 cmbs Friday knife Complete; beveled and serrated edges; Figure 7-20b

165AB-57 1 2 at 20 cmbs Hare biface Complete; Figure 7-20c 3 1 B 1 3 at 30 cmbs Native Ceramic Two Fragments (Appendix J) 6 1 A 1 3 20-30 cmbs Native Ceramic Nine Fragments (Appendix J) 167 BHT 24 . . backdirt Native Ceramic Two Fragments (Appendix J) 51 temporal affiliation may support the likelihood that Zone 3 in both blocks may date to the Late Archaic II.

Aboriginal Ceramics

Timothy K. Perttula analyzed thirteen prehistoric ce- ramic sherds recovered from Block 1, Units A and B, and BHT 24 (Appendix J). Eleven of the sherds were recovered in situ at approximately 30 cm bs, with the remainder encountered in the backdirt of BHT 24. The sherds are from two different vessels, a probable thick- walled jar and an incised carinated bowl. The prob- able jar sherds have finely crushed and burned bone temper inclusions, with an average wall thickness of approximately 8.8 mm. The incised carinated bowl sherds are tempered with bone and grog, with a sandy Figure 7-20. Chipped stone tools recovered from clay paste, and an average wall thickness of 8.15 mm. 41MM341. a) Scallorn arrow point; b) Friday knife; While the bowl sherds exhibit characteristics similar c) Hare biface. to Caddoan vessels, similar sherds have been reported from Post Oak Savanna sites in the Navasota and Brazos River basins (Perttula n.d.; Turpin and Car- Feature 1 penter 1994:17). The sherds are also reminiscent of Feature 1 is a refuse, or secondary, pit consisting of a Toyah Phase sherds from the Rowe Valley site (M. dark, midden-stained matrix with charcoal, charcoal- Tomka, personal communication, 2000). To ascertain stained clay, fired clay, and sparse to moderate mus- the manufacturing locale of the two different vessels, sel shell. The feature was encountered within the first Perttula recommends that a sherd from each vessel be arbitrary 10-cm (4-in.) level at 40 cm bs in Block 1, analyzed by instrumental neutron activation analysis XUD; however, the mechanical excavation to this (INAA) and petrographic analysis. The current INAA depth probably truncated an unknown amount of the and petrographic databases for Caddoan and Central upper portion of the feature. The extant remnant of Texas ceramics are sufficiently robust (see Perttula the feature exhibited a depth of approximately 8 cm 1999) to be able to determine the likely compositional (3 in.). The provenience of this feature, terminal at 48 origin of the two vessels. cm bs (19 in.), would suggest that it is associated with a somewhat later component than the shell lens, Fea- ture 3 (see Feature 3 description). Feature Descriptions Feature 2 A total of five cultural features were identified and Feature 2 was originally recorded during the TxDOT excavated at 41MM341. Three features were newly survey phase in BHT 15 as an apparent hearth feature recorded and two (Nos. 2 and 3) were previously re- (Ahr and Abbott 1999:13). Re-excavation of the back- corded during the survey phase. Two of the encoun- hoe trench allowed for the relocation of the feature. tered features (Nos. 4 and 5) can be classified as Upon manual excavation, however, Feature 2 appears hearths with associated fire-cracked rock, charcoal, to be a refuse, or secondary, pit feature associated with and other heat-altered remnants. Two features (Nos. 1 an unlined hearth, Feature 4 (see below) located ap- and 2) are interpreted as refuse pits. A single feature proximately 82 cm (32 in.) to the east. The feature oc- (No. 3) is expressed as a discrete shell lens. Table 7- curs at approximately 50 cm bs and extends to 15 lists the features encountered at 41MM341 by pro- approximately 75 cm bs. It appears that a majority of venience. More detailed descriptions of the individual the feature was removed with the initial survey back- features are discussed below. hoe trench excavation. The remnant portion of the 52 Table 7-15. Features encountered during manual excavation at 41MM341

Feature Block Unit Elevation Type 1 1 D 40-48 cmbs Unlined pit feature. 2 2 A, C 50-75 cmbs Unlined pit feature. 3 1 A, B, D 60-70 cmbs Freshwater mussel shell lens. 4 2 B, D 56-74 cmbs Unlined hearth feature. 5 1 B, D 75-82 cmbs Rock-lined hearth. feature is irregular in shape and measures 23 cm (9 in.) Block 1, XUB and XUD at 80 cm (31.5 in.) below wide and 115 cm (45 in.) long. The rich, midden-stained ground surface. Artifacts recovered from the feature soil of the refuse pit included burned clay, chipped stone, include fire-cracked rock, charcoal, bone, and sparse mussel shell, and small fragments of burned rock burned mussel shell fragments. The feature measures approximately 50 cm (20 in.) in diameter and appeared Feature 3 to exhibit a slight vertical concavity. Feature 3 was initially identified during the survey phase as a possible early stage lithic reduction sta- tion, with associated primary and secondary corticate Artifact Distribution flakes and an abundant concentration of mussel shell (Ahr and Abbott 1999:14). Subsequent to placement Tables 7-16 and 7-17 present the distribution of arti- of the excavation block (Block 1), it became apparent facts recovered from the two blocks of 41MM341. A that the feature was actually the westernmost portion look at the distribution of chipped stone, fire-cracked of a discrete shell lens, extending eastward under the rock (FCR), mussel shell, and animal bone in the upper elevated portion of State Highway 36. Within the ex- five levels of Block 1 shows a small peak in three of cavation block, the lens occurs from 60–70 cm bs. the four categories in Level 4 corresponding to Zone 1 The heaviest concentration of mussel shell occurs in (Table 7-16 and Figure 7-21). No equivalent peaks are XUB, with a peripheral scatter contained in the re- present in Block 2 (Table 7-17 and Figure 7-22). Zone maining test units of Block 1. 2 in Block 1 is characterized by peaks in chipped stone artifacts, FCR, animal bone, and mussel shell. The den- Feature 4 sities of all four categories peak in Level 8 of Block 1. Feature 4 is a large basin hearth and consists of a cir- In Block 2 chipped stone artifacts show only a small cular arrangement of relatively large (>10 cm) hearth- peak in Zone 2 (Figure 7-22). The bulk of the materials stones positioned around the edge of the basin. The from Zone 2 in Block 2 consist of FCR. A small peak in matrix within the basin is a dark, charcoal-stained animal bone is present in Level 6 of the block. Zone 3 material replete with fired clay nodules and charcoal in Block 1 consists of small peaks in chipped lithics, flecks. Sparse amounts of burned bone, chipped stone, FCR, and mussel shell in Level 10 followed by steadily and smaller burned rock were also recovered from the decreasing frequencies in all three categories in deeper interior of the basin. The feature was encountered at levels. Animal bone densities peak in Level 11 of the approximately 56 cm (22 in.) below ground surface. block. In Block 2, Zone 3 is characterized by higher The eastern wall of Block 2 bisects the feature, at densities of chipped lithics, mussel shell, and bone com- which point the diameter measures approximately 2 pared to Block 1. In addition to a peak in materials in m (6.6 ft.). Level 10 in both blocks, Block 2 also shows a smaller peak in chipped lithics, FCR, and mussel shell in Level Feature 5 15 (see Table 7-17 and Figure 7-22). Feature 5 is a rock-lined basin hearth and consists of a roughly circular concentration of fire-cracked rock with In general, these distribution patterns suggest some associated fired clay, charcoal, freshwater mussel shell, important spatial variation in activities carried out dur- and lithic debitage. The feature was encountered within ing the different occupations across the site. Toyah

53 Table 7-16. Vertical Distribution of Artifacts and Ecofacts in Block 1, 41MM341

Chipped Mussel Shell Mussel Shell Snail Shell Stone FCR Umbo Fragments Snail Shell Fragments Ceramic Bone Zones Level N N Wt N Wt Wt N Wt Wt N N Wt 1 1 0 0 0 0 0 1.32 1 0.06 0.09 0 0 0 1200000 0 00 0 041.34 1 3 5 0 0 2 1.9 7.49 0 0 0.04 9 1 0.02 1 4 21 0 0 1 19.1 2.32 0 0 0.69 2 25 29.87 1 5 3 5 5.78 2 6.2 6.59 10 5.5 0.81 0 10 22.31 2 6 9 5 3.06 14 54.4 59.22 205 14.8 3.83 0 6 0.87 2 7 68 24 234.25 61 265.4 195.17 246 80.5 17.19 0 37 7.27 2 8 174 229 1851.52 101 507 336.27 185 50.1 25.13 0 38 14.45 2 9 10 39 65.62 8 21.3 64.25 224 45.6 4.64 0 0 0 3 10 32 42 103.5 16 54 22.09 590 123.8 5.23 0 1 1 3 11 6 0 0 1 0.6 4.39 368 86.7 3.52 0 12 6.26 3 12 3 0 0 24 48.2 1.82 366 69.4 4.82 0 2 0.11 3 13 4 2 1.94 4 6.3 15.59 220 45.5 4.55 0 3 0.61 3 14 1 0 0 2 1.73 1.28 561 108.3 6.61 0 0 0 3 15 1 0 0 2 8.4 4.08 337 62.4 0.79 0 0 0 3 16 0 0 0 1 0.8 1.44 306 57.5 0.94 0 2 0.71 Totals 337 346 2265.67 239 995.33 723.32 3619 750.16 78.88 11 141 84.82

Table 7-17. Vertical Distribution of Artifacts and Ecofacts in Block 2, 41MM341

Chipped Mussel Shell Mussel Shell Snail Shell Stone FCR Umbo Fragments Snail Shell Fragments Bone Zone Level N N Wt N Wt Wt N Wt Wt N Wt 1 1 2 6 92.46 0 0 0 1 0.1 0 0 0 12100000 00 0 00 13000000 00 0 00 1 4 0 10 124.59 1 2.2 0 26 4.5 0.35 1 0.1 2 5 5 415 644.53 3 9.4 2098.79 8 3.5 1.88 29 34.58 2 6 21 2099 4418.37 7 34.3 10.18 29 21 2.89 6 1.99 2 7 9 481 3274.92 4 14.1 1.26 29 8 3.37 6 3.58 2 8 13 44 57.97 2 6.4 4.91 102 22.7 1.44 6 0.95 3 9 53 47 175.25 35 226.6 76.21 68 14.3 1.42 9 3.7 3 10 151 19 7.8 13 63.8 35.32 260 51.4 2.63 7 4.94 3 11 75 5 3.5 2 22.2 0.29 370 74.2 5.58 1 0.059 3 12 53 1 0.17 3 8.8 4.68 324 60.3 8.44 16 14.7 3 13 35 1 1.83 5 26.5 20.87 213 42.8 5.39 13 3.4 3 14 6 1 0.47 10 14.6 8.62 136 29 3.96 5 0.43 3 15 9 9 128.02 18 43.5 22.34 11 2.1 0.83 3 0.07 3 16 6 3 11.65 7 9.2 13.68 172 36 0.65 1 0.03 Total 439 3141 8941.53 110 481.6 2297.15 1749 369.9 38.83 103 68.529

54   ]RQHÃ   OV  H Y  H ]RQHÃ Ã/ Q  R WL  D Y  D  [F (    ]RQHÃ   

        1XPEHUÃRIÃ&KLSSHGÃ6WRQHÃ,WHPV

Figure 7-21. Distribution of chipped stone items by zone in Block 1 at 41MM341.

  ]RQHÃ   OV  H Y  H ]RQHÃ Ã/  Q R WL  D Y  D  [F (    ]RQHÃ   

        1XPEHUÃRIÃ&KLSSHGÃ6WRQHÃ,WHPV

Figure 7-22. Distribution of chipped stone items by zone in Block 2 at 41MM341.

55 Phase occupation debris seems to be concentrated in zones strengthen the conclusion that these zones are the vicinity of Block 1. During the Austin Phase, the distinct archaeological components. The breakdown eastern edge of the site may have been used as a gen- of cortex categories within the debitage sample from eralized discard area, while the western margin of Block 1 indicates that the percentage of entirely the site may have been reserved for cooking activi- decorticate debitage increases with increasing depth ties associated with maintained facilities. Judging of the zones (Table 7-18). That is, Zone 1 has the lowest from the differences in material densities between the percentage of decorticate debitage (66 percent), while two blocks, differences in the use of space may also Zone 3 has the highest (74 percent). Roughly the same have existed during the Late Archaic II occupation of pattern is noted in the debitage sample from Block 2 Zone 3. (Table 7-19). Only three flakes were encountered in Zone 1 of the block, but in Zones 2 and 3, where sample sizes are adequate (n=47 and n=383, respectively), the Debitage Analysis increased percentage of decorticate debitage is clearly evident with increasing depth. All unmodified debitage recovered from the two blocks at 41MM341 has been categorized in terms of Overall, the percentage of decorticate debitage in the corticate (i.e., primary and secondary) and decorti- two blocks is not unlike those noted in the 41MM340 cate (tertiary) specimens. A comparison of relative debitage samples. At this point, insufficient data ex- proportions of these two categories can identify dif- ists to determine whether the trends noted in the two ferences in stages of reduction either between or blocks of 41MM341 are due to behavioral differences within blocks and between the three cultural zones regarding lithic reduction strategies or the impacts of of the site. depositional, site formation, processes. If due to be- havioral factors, these patterns would reflect an intrigu- Based on the distribution of the major artifact ing trend through time associated with the degree of categories in Blocks 1 and 2, three zones were reduction of tools manufactured and/or rejuvenated on distinguished. Radiocarbon assays and site. If depositional in nature, the patterns may have chronologically sensitive artifacts found within these some important impacts on the artifact assemblages associated with each zone.

Table 7-18. Percentage of cortex categories in Block 1, 41MM341

Zone Corticate Decorticate 13466 23070 32674

Table 7-19. Percentage of cortex categories in Block 2, 41MM341

Zone Corticate Decorticate 16733 23268 32773

56 Vertebrate Faunal Material evolved in flowing waters and are most often associ- ated with gravel, stable sand, and mud substrates in Barbara A. Meissner, Center for Archaeological Re- permanent-water conditions. These data suggest the search, The University of Texas at San Antonio pro- site was formerly in relative close proximity to a pe- vided analyses of vertebrate faunal material recovered rennial stream; possibly a former meander of the Little from the two excavation blocks at 41MM341 (Ap- River. pendix C). According to her report, the predominantly fragmented bone is generally in moderate to good con- dition, with four genera of mammals and three genera Macrobotanical of reptile identified. The majority of the unidentified bone appeared to be from animals of the deer/ante- J. Philip Dering, Associate Director for Macrofossil lope size category. Additionally, highly fragmented, Analyses, Paleoethnobotany Laboratory, Texas A&M unidentifiable specimens of bird and turtle were University performed macrobotanical analyses of a present. There is little evidence of extensive atmo- single flotation sample recovered from the shell lens spheric or chemical weathering, suggesting clayey, flu- encountered within Block 1. According to his report vial deposits buried the material shortly after initial (Appendix E), the archaeobotanical assemblage, both fragmentation and subsequent deposition. Overall, a in quantity and size of the material, is very reduced. relatively small amount of faunal material was encoun- Of the sample submitted for analysis, only 6 very small tered within this otherwise rich, cultural deposit. In (<2 mm) thin nutshell fragments of an indeterminate addition, a majority of the bone encountered is highly species (cf. pecan) were encountered. Due to the over- fragmented, suggesting the material was heavily pro- all paucity of botanical remains exhibited in the gen- cessed. Further analyses may provide a clearer inter- eral level samples submitted, future analyses should pretation of the relative paucity of remains and the be directed toward samples recovered from heating extent to which the remains present were processed. features and the samples collected from radiocarbon dating, both of which are conducive to the preserva- tion of botanical materials and may yield larger iden- Mussel Shell tifiable specimens.

Robert G. Howells conducted analyses of freshwater mussel specimens recovered from the shell lens en- Pollen and Phytolith countered in Block 1 at 41MM341 (Appendix D). Forty-nine specimens selected to show the variability Analyses of pollen and phytolith samples extracted in species present in Block 1 were submitted for analy- from the shell lens present in Block 1 were performed sis. According to his report, all specimens were in very by John G. Jones, Associate Director for Pollen Analy- good condition, with four species identified, repre- ses, Palynology Laboratory, Texas A&M University senting the Family Unionidae. Threeridge (Amblema (Appendix F). Pollen and phytolith specimens were plicata) represented roughly one-third of the sample; noted within the samples submitted, however, the over- smooth pimpleback (Quadrula houstonensis) repre- all state of preservation was poor. The pollen concen- sented nearly 60 percent of the sample; with southern tration value for the single cultural zone is 1,694 mapleleaf (Quadrula apiculata) and Tampico grains/ml. Environmental factors such as cyclic wet- pearlymussel (Cyrtonaias tampicoensis) comprising ting and drying in combination with the acid soils of the fractional remainder. Comparison of this assem- the region are not conducive to the preservation of blage with modern collections from the Central Brazos pollen or phytoliths, and, as such, the overall yield, or River Drainage mirrors the regional climatic trend to- concentration value of the individual samples do not ward a more arid environment exhibited by the change lend themselves to reliable interpretations. in mussel population at 41MM340. Similarly, all the taxa represented in the archaeological assemblage

57 Land Snail Shell exemplifies the higher probability of recovering dia- tom populations from fired clay nodules that may tend Richard W. Fullington conducted analyses of two gas- to seal in and preserve silica diatoms preventing them tropod samples extracted from the shell lens in Block from entering back into solution in these types of silica- 1 (Appendix G). The samples consisted of specimens poor soils. collected from floated light and heavy fractions and those specimens obtained from ¼-inch water screen- ing. According to his report, all specimens were in Summary excellent condition, with eleven species identified, rep- resenting eight families. When the sample of gastro- Our limited testing in the two blocks at 41MM341 pods is divided into their respective preferred habitat suggests that three distinct cultural zones may be types, it becomes evident that species preferring a present. These zones are defined by dramatically dif- moist, densely wooded setting greatly outnumber ferent densities of chipped stone and fire-cracked rock. (n=168; 86 percent) those inhabiting open or sparsely These differences are supported by different types of wooded settings (n=29; 14 percent). The species rep- diagnostic artifacts and different frequencies of fea- resented suggest similar environmental conditions to tures. Radiocarbon dates support these distinctions. A those encountered in Zone 4 at 41MM340. A notable radiocarbon date from Zone 1 suggests an occupation exception, the presence of Anguispira alternata, sug- data of around A.D. 1407–1440, a period that corre- gests the presence of exposed, wooded limestone sponds with a Toyah Phase component. Zone 2 has outcroppings within the vicinity. The gastropod shells two radiocarbon dates, one that spans A.D. 1020–1160 examined appear to be deposited in situ and not from and one that has a range of A.D. 685–780. These dates fluvial transport. The relatively low number of spe- suggest that this zone spans the Austin Phase. Finally, cies present and the occurrence of only one species of while we lack any radiocarbon dates for the Zone 3 aquatic gastropod qualify this. Additionally, the pres- material, it is clearly manifested in both blocks by large ence of the predatory snail, Haplotrema concavum, is increases in chipped stone. The recovery of several indicative of a stable and long-term gastropod com- diagnostic artifacts suggests that this occupation may munity, occurring only within large populations of land date to the Late Archaic II period. snails. In comparison with 41MM340, there appears to be little difference in the frequency of specimens encountered among the cultural zones. These data sug- gest that the single cultural zone represents a relatively short time span with a low snail population density.

Diatom

Barbara Winsborough, of Winsborough Consulting, conducted analyses on two diatom samples collected from site 41MM341 (Appendix H). One sample was collected from the sediment matrix comprising the shell lens exhibited in Block 1. A second sample, con- sisting of fired clay nodules, was collected from a hearth feature extending just beneath the shell lens in the same block. While there appeared to be sufficient freshwater mussel shell present in proximity to a hearth feature to mirror the diatom assemblage recovered from 41MM340, evidence of diatoms was absent from each of these samples. This differential preservation

58 Chapter 8: Recommendations

Site Eligibility The separation of the site into northern and southern sections is justified by the presence of an apparent former slough bisecting the midsection of the survey- 41MM340 phase-delimited site bounds. Results of auger borings identified an approximately 20-m wide sterile corridor Site 41MM340 is interpreted as a prehistoric multi- separating the two sections; a distance typically re- component open campsite located above and within garded as grounds for determination of separate sites, various alluvial straths deposited atop a sandy pointbar especially with the probability of a dynamic stream along the left descending bankline of a probable former dissecting the former landscape. It is equally possible, meander of the Little River. In support of this inter- however, that the single component exhibited in the pretation, core boring results from the 1930s suggest northern section of the site was at one time a portion that a relatively wide channel cut through the underly- of at least one of the components of the southern sec- ing strata at depths over 10 m below current ground tion of the site. Based solely upon the radiocarbon surface within approximately 30 m of the site. Nine dates, the contemporaneity of the feature in Block 1 different zones are visually identifiable in the two ex- with Zone 6 in Blocks 2 and 3 is evident. In addition, cavation blocks in the southern portion of the site, as while Block 1 was excavated in arbitrary 10-cm lev- expressed within the rights-of-way. els, the distribution of chipped stone by level clearly indicates two, and possibly three, peaks that may re- Within each block and between blocks there is signifi- flect similar, though less visible, occupation zones. cant variation in thickness of all zones. The degree to which they are identifiable and separable during exca- Based on artifact content and geomorphic character- vation varies. The dating of the upper and lower por- istics all cultural zones could be visually identified and tions of the four cultural zones with the highest artifact may be correlated across both southern blocks (Blocks density within and across the two blocks in the south- 2 and 3) at 41MM340. In general, the zones share ern portion of the site indicates that, while accumula- similar proveniences below ground surface. Nonethe- tion rates vary, several zones have accumulated so less, even within the same block vertical variation due rapidly that the dates are statistically distinguishable. to the undulating surfaces, and truncation and merging Furthermore, the only zones that can be shown to be of some zones made it difficult, at times, to follow a statistically different from each other, in terms of dates, zone across the entire block. Zone 2 is expressed as a are Zone 2 and Zone 8, the upper most and lowest shell midden throughout Block 3 and the same zone cultural zones. Nonetheless, the relatively large series may be manifested in the western portions of Block 2. of radiocarbon assays run on material from the site Zone 4 is expressed as a fire-cracked rock scatter have established that the approximately 130-cm thick throughout both blocks. A similar fire-cracked rock vertical column probably represents less than approxi- scatter with a greater concentration of freshwater mately 1,100 years of prehistoric occupation. Overall, mussel shell comprises Zone 6. Zone 8 is expressed the cultural material recovered from this portion of the as a scattering of fire-cracked rock with a relative site is rich and contains archaeological assemblages abundance of lithics and a sparse scattering of fresh- highly relevant to the investigation of Late Archaic I water mussel shell. Zones 3, 5, and 7 are, in turn, char- and Late Archaic II Period cultural adaptations within acterized by lower densities of chipped stone, this region. fire-cracked rocks, mussel shell, and a lack of fea-

59 tures. The analysis of the radiocarbon dates, discussed the Little River. Auger boring results have indicated a in Chapter 7, indicates that in Block 2, the upper por- relatively large single areal concentration of materials tion of Zone 2 is indistinguishable from the lower part within the rights-of-way. These limits fall between two of the same zone, as well as from the upper and lower extant sloughs and, according to the 1930s core portions of Zone 8. In Block 3, the upper and lower borings, immediately south of a former large stream, portions of Zone 2 are statistically different from Zones probably a meander of the Little River. A single cul- 4 and 6, but these two zones are indistinguishable from tural occupation zone is evident in both Blocks 1 and each other. Finally, Zone 8 is statistically different from 2, as expressed within the rights-of-way. In addition, a all overlying zones. discrete later component is apparent in the upper por- tion of Block 1, and a possible discrete earlier compo- These patterns suggest that cross-dating of the zones nent is suggested in the lower portion of Block 2. between the two blocks of the site might not be a Further, the three radiocarbon assays have suggested simple matter. They also suggest that some degree of a relatively clear resolution of the chronology within vertical mixing may have occurred between the zones. the vertical column for the approximately 750 to 800 The results of the extensive amino acid racemization years of prehistoric occupation. The cultural materials analyses conducted on snails and mussel shell from recovered from the site moderate and contains an ar- the various cultural zones also hints at this possibility. chaeological assemblage complementary to the inter- The analysis of A/I ratios for a sample of two species pretation of the later part of the Late Archaic II and of snail shells (Polygyra mooreana and Praticolella both phases of the Late Prehistoric Period within this berlandieriana) and two species of mussel shells region. (Tritogonia verrucosa and Amblema plicata) also suggests some movement of material between zones. The upper, discrete component identified in Block 1, consisting of small pit features and aboriginal pottery, In summary, it is the opinion of the authors that, in was probably truncated by the mechanical stripping of general, site 41MM340 contains several clearly distin- ¾ of the block prior to manual excavation. The cor- guishable cultural zones with high research potential. rected and calibrated radiocarbon date, with a 1 sigma Furthermore, testing has demonstrated that isolated seg- range of A.D. 1407 to 1440, combined with the pres- ments of the site may potentially contain a number of ence of ceramics, suggests a possible Toyah Phase visually and analytically separable components accu- component (Prewitt 1981:83), worthy of further inves- mulated over short time periods. Therefore, it is the tigation. The underlying shell lens within the same block recommendation of CAR that site 41MM340 be nomi- also provides a convenient correlation between the ra- nated for inclusion in the National Register of Historic diocarbon date and the Scallorn arrow point, suggest- Places (NRHP). Specifically, the degree of fine- ing the earlier Austin Phase of the Late Prehistoric grained temporal resolution afforded by the deposits Period (Prewitt 1981:82). The large hearth feature of the site, as well as their potential to allow the recon- encountered in Block 2 provided a radiocarbon date struction of paleoclimatic conditions and the investiga- with a 1 sigma range of A.D. 1020 to 1160 and was tion of changes in human adaptation over short time underlain by a Friday biface, a contemporaneous knife periods rarely seen in other archaeological sites in the temporally contained within the Austin Phase (Prewitt state, strongly argues for recommending the site to 1981:82). In addition to these feature-rich zones, the the NRHP (36 CFR 60.4 Criterion (d)). underlying artifact peak at 100–110 cm bs may date to the Late Archaic II and may also be worthy of further analysis. 41MM341 Through synthesis of the radiocarbon dates and analy- Site 41MM341 is similarly interpreted as a prehistoric ses of the few diagnostic artifacts, the individual com- multicomponent open campsite located above and ponents of 41MM341 suggest a moderate to high level within various alluvial straths deposited along the right of temporal integrity, with minimal bioturbation evident. descending bankline of a probable former meander of However, the reconstruction of the paleoenvironment

60 for the site will be difficult, at best, without the aid of typical climatic indicators. Special analyses conducted for the presence and identification of diatom, pollen, phytolith, and macrobotanicals have all returned some- what limited results. The strategy employed for pre- liminary analyses focused upon the cultural zones, or occupation surfaces themselves, rather than the indi- vidual features encountered within the zones. While the radiocarbon dates have provided a relatively clear and tight chronology for the site, a different strategy specifically addressing investigation of the features contained within the cultural zones must be employed for interpretation of the site, climatically as well as temporally.

In conclusion, it is the opinion of the authors that 41MM341 exhibits such qualities and overall temporal integrity to be eligible for inclusion in the National Reg- ister of Historic Places (NRHP). Specifically, evalua- tive consideration in regard to temporal integrity and the ability to provide significant data for reconstruction of past environments provide for recommending NRHP eligibility of 41MM341 (36 CFR 60.4 Criterion (d)).

61

Chapter 9: Data Recovery Plan

Data Types and Research Issues into valley-proximate uplands may have occurred. Following another decrease between 800–500 BP, The following data types have been recovered from moisture levels increase substantially resulting in the the test excavations at 41MM340 and 41MM341: return of woodlands into modern times. chipped lithics, burned rock, mussel shell, snail shell, large quantities of charred macrobotanical remains, One of the most important features of these oscilla- diatoms and small quantities of pollen and phytoliths. tions is their impact upon the local and regional The general characteristics of these data types and the resource base available to hunter-gatherers. Through- research questions that can be addressed with them out prehistory plant resources have always played a are discussed below. significant role in hunter-gatherer subsistence. How- ever, with the exception of the origins of food production, archaeologically in Texas the most vis- Paleoenvironmental Data ible and best documented shifts in prehistoric hunter- gatherer subsistence strategies appear to be correlated The review of the existing data and reconstructions of with the increase of hunting and the role of animal the paleoenvironmental conditions characterizing the protein in prehistoric diets (e.g., Toyah). It is perhaps last 3,000 year period of the occupations of 41MM340 because of the impact of such shifts on land use, and 41MM341 indicates that this period was charac- subsistence strategies, and technological organization terized by oscillations between moist and dry periods that these shifts are more recognizable in the of various duration. Based on the microfauna from archaeological record. Hall’s Cave a wet period characterizes the period be- tween about 3200–1000 BP, with moisture levels reach- The changes in paleoclimatic conditions outlined ing their peak between about 2500–1300 BP (Collins above affected the range and productivity of prairie 1995). Based on data from Boriack and Weakly Bogs, and woodland vegetation exploited by prehistoric Bousman (1998) reconstructed the estimated arboreal hunter-gatherers. In addition, they also affected the canopy cover throughout the Late Holocene. This range and density of animal species available for hu- reconstruction shows a period of increased arboreal man consumption. Most importantly, changes in ef- canopy cover –typically associated with increased fective moisture levels have impacted the productivity effective moisture levels– between 3500–3000 BP. of the grassland communities on the Blackland Prai- Drier conditions characterize the period between rie and the neighboring Post Oak Savanna. 3000–2200 BP with moisture levels bottoming some- time around 2500 BP. This period is characterized by Although some disagreement exists, greater effective grassland vegetation communities. Effective moisture moisture levels appear to have favored the return of levels appear to increase thereafter reaching a peak large numbers of bison into Central Texas and onto between 1700–1900 BP, and correspond to an increase the Blackland Prairie. Reconstructions of regional in the arboreal canopy. Between 1600–1300 BP, effec- populations (Dillehay 1974; Collins 1995) indicate tive moisture levels appear to drop again and grass- that bison were present in Texas during the middle to lands appear to increase in extent at the expense of late portion of the Late Archaic until approximately woodlands. Although the Boriack and Weakly bog data 1600 BP. In contrast, bison remains become scarce indicate a slight increase in effective moisture levels during the very late portion of the Late Archaic when between about 1200–800 BP, the associated vegeta- Darl projectile points are in use and this scarcity con- tion communities of the time period remain dominated tinues throughout the Austin Phase of the Late by grasslands although some expansion of woodlands Prehistoric period. The Late Archaic period of bison

63 presence appears to correlate well with higher largest of the snail sample from Zone 8. Interestingly, effective moisture levels within the region. Similarly, a single specimen of Haplotrema concavum, a the extended period of bison scarcity that appears to predatory species indicative of stable long-term be present during the Austin Phase of the Late Prehis- gastropod populations, is present in Zone 6. Its toric period also appears to correlate with low levels presence suggests that snail populations remained large of effective moisture as evidenced from the Patschke throughout the time Zones 6 and 8 were occupied. The Bog core (Camper 1991). snail population from 41MM341, occupied some 850 years after the terminal occupation of 41MM340, is More detailed knowledge regarding the paleoclimatic dominated by species preferring dense woody cover. conditions characterizing the 700–1000 year period This shift compared to the Zone 2 population at reflected in the occupation zones at 41MM340 and 41MM340 suggests an additional change in the broader 2500 years that include the last occupa- vegetation communities. tion of 41MM341 will allow greater precision in mod- eling the changes in the nature and structure of the It is likely that the composition of the snail popula- resources that were potentially available in the imme- tions by preferred habitat type reflect both macro- diate vicinity of the sites and in the neighboring veg- regional climatological conditions and changes in etation communities. plant community composition and structure in the immediate vicinity of the two sites. The presence of One research direction that could aid in the recon- Praticolella berlandieriana in combination with the struction of paleoenvironmental conditions in the vi- large Rabdotus population may reflect the initial stages cinity of the sites and on the floodplain in general is a of stabilization of the floodplain (i.e., decreased systematic documentation of flood frequencies asso- frequencies of flooding), the early stages of plant ciated with the Little River during the occupation of invasion, and the filling in of this newly available the two sites. Potentially the best location to pursue niche by a large colonizing snail population. The this is in either of the two sloughs found in the vicin- appearance of the predatory snail species may be in ity of the sites. In the same vein, a thorough compre- response to the exploding colonizing population sizes hension of the history of the Little River floodplain within this new niche. The decreased proportion of where sites 41MM340 and 41MM341 are located may open grassy/sparsely wooded habitat species in Zone provide a better understanding. 6, and the dominance of densely wooded habitat species in Zone 4 may reflect the gradual emergence Although the snail samples analyzed by Fullington of a stable floodplain with a riparian woodlands (Appendix G) represent relatively small proportions vegetation. The shift back to exclusively open grass- of the available sample, and only a small proportion land/sparsely wooded habitat species in Zone 2 may consists of the “micro” snails derived from the light be a reflection of regional scale decreases in effective and heavy fractions acquired from flotation, some in- moisture, and the resulting thinning out of the woody triguing differences in population composition are component of the vegetation. The 41MM341 snail notable through time. For instance, grouping the snail population may indicate yet another oscillation in samples from 41MM340 by the preferred habitat types climate conditions or vegetation succession on the provided by Fullington indicates a trend from a domi- Little River floodplain. nance of snail populations by open grassy/sparsely wooded species (98.4 and 63.9 percent in Zones 8 and Although this preliminary reconstruction of 6, respectively), to a shift to moist densely wooded paleoclimatic contexts is based on a very small sample, habitat species in Zone 4 (84 percent). The Zone 2 it is clear that a systematic study of fully representa- snail population only contains species tolerant of open tive samples from both 41MM340 and 41MM341 can grassy/sparsely wooded habitats. In addition, two provide valuable information both towards an specimens of Praticolella berlandieriana, a specie that assessment of the model described above, as well as a prefers benched terrace type settings with early more detailed understanding of floodplain dynamics succession emergent vegetation, was present in the within the vicinity of the two sites.

64 A number of specific research issues relevant to the 80 light and heavy fraction samples recovered from reconstruction of paleoenvironmental conditions can the two sites. be defined based on the preceding discussion. They include, but are not limited to, the following: Preliminary analysis of the vertebrate bone from the two sites suggests that the bulk of bone represents deer 1) What are the paleoenvironmental conditions and antelope-sized animals. In addition, based on cor- characterizing the roughly 3,000 year period tical thickness, only two small fragments may be from encompassing the spans of occupation of the bison/cow-sized animals. The diversity of vertebrate two sites? taxa (fish, turtle, rabbit, possum, raccoon, bird, ante- lope/deer), and the presence of moderate numbers of 2) Is there a shift from predominantly grassland mussel shells suggests a wide diet breadth. The addi- to floodplain forest conditions during the early tion of pecan nuts and likely other plant species to the portion of the Late Archaic I as reflected at diet further extends the diversity of diet. The small 41MM340? size of the mammalian bone may also be considered 3) What is the history of the Little River flood- as a possible indicator of some degree of subsistence plain in the vicinity of the two sites and how stress. Both aspects of the subsistence remains are does it relate to the occupation histories of somewhat surprising given the likelihood that bison the sites? populations were present and exploited throughout Central Texas and presumably on the Blackland Prai- 4) What can we learn from the geomorphological rie particularly during the approximately 700 years history of the Little River floodplain that can encompassed by the 41MM340 occupation zones and aid in future assessments of the likelihood of during the last occupation zone (Toyah Phase) present preservation of historic properties in similar at 41MM341. contexts? 5) What shifts, if any, occurred in the Prairie/ The scarcity or perhaps entire lack of bison remains Post Oak Savanna boundary during the 3,000 at the two sites is surprising given the generalized im- years the two sites were occupied? pression gained from the Late Archaic Central Texas Subsistence Data archaeological record. It is also surprising given the generalized assumption from optimal foraging that if bison was ever encountered by local populations it Almost 1,500 pieces of bone have been recovered in would have been pursued given its high ranking. Nu- addition to about 6,770 mussel shell umbos and over merous factors may account for the composition of 22,500 whole snail shells from 41MM340. In addi- the faunal assemblages recovered from the two sites. tion, three out of four macrobotanical samples yielded They include, but are not limited to, the following: pecan and/or hickory nutshell fragments. Site 41MM341 yielded 244 pieces of bone, 349 mussel 1) The effects of transportation costs on skel- shell umbos, and over 5,300 whole snail shells. The etal element representation; single macrobotanical sample produced nutshell frag- ments. The condition of all of the faunal remains is 2) The impact of food processing techniques on exceptional, indicating that differential preservation skeletal elements; conditions are not the primary factors responsible for 3) Site functional considerations; the composition of these samples. The condition of the remains also indicates that depositional and 4) Site structural considerations; preservation conditions are favorable for the recov- 5) The absence/scarcity of bison within the re- ery of additional materials in these classes. The rate gion; and of recovery of nut shell remains (3 out of 5 samples; 60 percent), suggests that edible plant remains (e.g., 6) Regional land use strategies that do not ex- pecan, hickory) may be common in the 84 14C samples tend out onto the Blackland Prairie and do containing the largest macrobotanical remains and the not rely on bison procurement.

65 A number of specific research questions can be The degree to which collections of lithic debitage and generated relating to these factors and potential sub- stone tools can be used to answer these issues will sistence strategies employed by site inhabitants. They depend in part on how clearly distinct components include, but are not limited to, the following: can be identified within sites. This, in turn, will depend on the dating of the components, the 1) Were the inhabitants of the two sites bison distribution of diagnostic artifacts, and information hunters? derived from snail and mussel shell racemization on disturbances and deposition rates by component. 2) Do the isolable components present at the site Nonetheless, the integrity, and by extension the represent “out-of-hunting season” occupations research potential of each component within the sites when groups subsisted on lower ranked and each site as a whole, should be seen as highly resources? dependent on the research questions and issues being 3) What was the proportional contribution of asked or pursued. That is, all sites have the potential terrestrial fauna, flora, and freshwater to contribute scientific knowledge, given selected resources to the diet? research issues and questions. The goal of the more comprehensive research design, to be developed at a 4) Were there oscillations in the proportional later date, is to define specific and detailed research contribution of these resources to the diet over questions and data types relevant to addressing them. time (i.e., between the cultural zones present at the two sites)? For some questions, sites situated at the high end of the 5) Does the fragmentation of the animal bone integrity continuum are critical, but for other questions, reflect dietary stresses or the impact of sites that traditionally are viewed as having low transportation limitations? integrity, are critical. Unmodified lithic debitage is helpful in informing about the nature of the tool manu- Chipped Lithics facture activities carried out at these sites (e.g., early, middle, and late reduction, tool rejuvenation, etc.), the A total of 9,545 chipped lithic artifacts was recovered reduction strategies employed (bifacial, core reduction/ from 41MM340. Site 41MM341 yielded 776 chipped blade-flake production), and the raw material procure- lithics. The large majority of these consist of unmodi- ment and transport strategies (e.g., local vs. non-local fied lithic debitage. Only nine projectile points have raw materials, core preparation activities, etc.) used by been recovered and the preliminary sort of the lithic the site inhabitants. The presence of abundant, good artifacts resulted in the identification of only nine non- quality lithic resources within the current channel of debitage lithic artifacts consisting of a Friday knife, a the Little River suggests that no matter where the bifacial knife fragment, a biface manufacture failure, prehistoric stream channel may have been, it served as two Erath bifacial adzes, a cobble tool, a nutting stone, a reliable source of raw materials. Therefore, it is and a hammerstone. expected that the majority of the lithic materials found on site would be of local origin. Nonetheless, given The chipped lithic artifacts recovered during testing that finished tools and blanks manufactured in other and those to be recovered during data recovery will parts of the annual range would have been brought onto be critical in investigating three main research issues the site, it is also expected that a small proportion of and directions: lithic debitage, of the smaller size range, would be of non-local origin. Whether these specimens can be iden- 1) Range of activities carried out at the site; tified would depend on the similarities in raw materials to those present within the gravel bars of the Little River. 2) The technological organization of the hunter- gatherers, this also reflects on larger scale off- The composition of the tool assemblages recovered site land use organization; and from the site should help identify the range of activi- 3) Regional and cultural relationships between ties carried out by the site occupants, as well as con- Central and East Texas. tribute to the understanding of overall site function

66 within the broad land use system. For instance, the Based on the previous discussion, a series of specific preponderance of manufacture failed vs. use broken research issues can be suggested: tools should suggest that lithic activities were oriented towards some degree of gearing up in anticipation of 1) Do the lithic assemblages recovered from the future needs. In contrast, the predominance of use isolable components of the two sites reflect failed specimens should be more characteristic of broad ranges of activities characteristic of retooling activities typically associated with tool long-term occupations, or a narrow range of maintenance carried out by foragers. Similarly, the activities more likely to have occurred on proportion of formal vs. expedient and minimally special activity sites? retouched tool forms may be indicative of degree of mobility within the annual cycle or the degree 2) Are there changes in assemblage composition of processing requirements experienced by the between the isolable cultural zones and do tool users. they reflect technological adaptations to shifting resource bases? One of the emerging research questions in Central/ 3) Do the assemblages recovered from the East Texas archaeological research has been the distinct cultural zones reflect contemporary definition of the place of east-central Texas in the land occupations by East Texas and Central Texas use strategies of East and Central Texas hunter- cultural groups? gatherer populations. Results of archaeological 4) Are there distinct artifact classes in the research in the region have for a long time suggested assemblages from the two sites that can be that the material culture present in the area showed a used to reconstruct territorial ranges, and great deal of affinity to Central Texas in terms of what were the territorial ranges of the sites’ projectile point types. These patterns are also inhabitants? confirmed by the distribution of projectile point types across these regions (Prewitt 1995). In addition, while 5) Are the ceramics recovered from 41MM341 a large proportion of the Late Archaic points and tool more typical of Toyah Phase specimens or forms in Central Texas are manufactured of locally do they belong with other East Texas available fine-grained cherts, some Late Archaic tool cultural units? forms characteristic of East Texas are consistently 6) Given the fine-grained control afforded by the made on fine-grained quartzites and/or petrified wood isolable cultural zones, is it possible to refine (e.g., Gary projectile points, Wooden points, the projectile point sequences of the Late Harvey bifaces, Bronson bifaces). Archaic I and II as they pertain to the Central Texas/East Texas archaeological regions? The breakdown of tools and debitage into raw mate- rial types may provide one gauge to measure the presence of cultural traits that are typically considered as deriving from one or another of these regions. This information may, in turn, reflect on whether the site is employed by hunter-gatherer groups from the Edwards Plateau or East Texas groups moving onto the Blackland Prairie or perhaps a mixture of both reflecting some type of communal activity relationships.

67 Methodology exterior test units, would not be explored, this plan would provide for minimal mitigation to investigate 41MM340 at least a portion of the southern section. Data Extensive excavation of auger borings revealed two recovery would thus be expected to multiply the distinct areal concentrations of cultural material, sepa- results of the testing efforts approximately six-fold. rating the site into northern and southern sections by an With the terminal excavation depth of 1.6 m, data re- approximately 20-m wide east-west corridor. The north- covery volume would equal 3,000 m3. ern section of the site contained a relatively thin, in- consistent zone of sparse scattering of cultural material. 2) Scenario 1 plus an additional 25 m2 Conversely, the southern section revealed a relatively exploratory grid. abundant amount of cultural material, with at least two a) Based upon the west wall profile of BHT B, it stratified components discernible within the auger is apparent that the most abundant continuance of borings. Moreover, excavation of test units in the south- the stratified components occurs to the west of ern section revealed two additional stratified compo- the tested portion of the site. Unfortunately, this nents occurring below the depths attainable by the auger, area is mantled by a substantial artificial berm, for a total of four stratified cultural deposits. which curtailed exploratory efforts during the test- ing phase. The berm rises 1.5 m above the current Accordingly, recommendation is made for further in- ground surface and has a cross-sectional width of vestigations to concentrate only upon the southern approximately 25 m at the base. While it is pos- section of 41MM340 in view of the following: the sible that subsurface disturbance occurred outside four distinct stratified components, good stratigraphic of the proposed ROW for the construction of the correlation between the test units, and the relatively berm, it is unlikely that the ground surface high artifact density. currently underneath the berm (i.e., within the proposed ROW) was impacted for its construc- Within the area of potential effect, the southern sec- tion. Thus, under this scenario, an approximately tion of the site comprises approximately 450 m2. Based 10-m wide swath of the berm would be removed upon the extensive auger data recorded during the test- for exploratory auger boring similar to that con- ing phase, a single areal concentration of stratified ducted during the testing phase. The results of the deposits has been delimited at approximately 150 m2. auger borings would aid the identification of the Further, as evidenced in the southern test excavation continuance of the cultural strata, if any, that units, the uppermost aspect of the latest cultural zone would condition the placement of the additional occurs at approximately 50–60 cm bs. In accordance 25 m2 exploratory grid. with the scope of work implemented for the testing phase, the upper 40 cm of recent alluvial fill can be b) In the event that exploratory efforts fail to re- mechanically removed without impact to the young- veal a continuance of the cultural zones west of est component. the site, then the additional 25 m2 exploratory grid would be absorbed by the main grid; providing a Focusing upon the areal concentration of stratified total main grid area of 75 m2. This area would deposits, the three following scenarios are proposed suffice to mitigate an adequate portion of the for (1) minimum, (2) adequate, and (3) maximum delimited site bounds, as well as a majority of the feasible data recovery: known areal concentration of stratified components. Both options (‘a’ and ‘b’) under 1) Excavation of an approximately 50 m2 grid, Scenario 2 would expect to yield approximately encompassing the southern test units. nine times the amount of material recovered during the test excavations. With the terminal This strategy would target the heart of the known excavation depth of 1.6 m, data recovery volume areal concentration of stratified components. While would equal 6,750 m3. the ephemeral boundaries, indicated by the continu- ance of the stratified zones within wall profiles of the

68 3) Combination of Scenario 2-a and 2-b. Focusing upon the overall site structure, the following scenario is proposed for data recovery This strategy would provide for exploratory efforts at 41MM341: west of the site and near complete mitigation of the areal concentration of stratified components. Again, 1) Excavation of an approximately 100 m2 if exploratory efforts fail, then the additional 25 m2 exploratory grid would be absorbed by the main grid; grid, encompassing the test units. providing a total main grid area of 100 m2. This large This strategy would target excavation of the area grid would adequately mitigate the areal concentra- contained within the 15 m separating the two portions tion of stratified cultural material in the southern of the site. With the dearth of cultural material portion of the site. Data recovery would thus be recovered during the testing phase, the focus of the expected to multiply the results of the testing efforts data recovery would be location of possible indica- approximately thirteen-fold. With the terminal exca- tors of site structure and formation. Thus, the vation depth of 1.6 m, data recovery volume would uppermost component, displaced by the use of heavy equal 12,000 m3. machinery during test excavations, would be further delineated, the areal extent of the southern portion 41MM341 of the midden would be determined, and the Extensive excavation of the auger borings revealed interrelation of the hearth and pit features and the shell a delimited areal extent of cultural material midden would be investigated. Data recovery would approximately 875 m2 as expressed within the rights- thus be expected to multiply the results of the testing of-way. Throughout, the site exhibited a relatively thin, efforts approximately twelve-fold. With the terminal inconsistent zone of sparse scattering of cultural excavation depth of 1.6 m, data recovery volume material. Excavation of test units corroborate the would equal 16,000 m3. results of the auger borings, with the identification of a thin midden in the eastern portion and a small complex of likely associated hearth and pit features in the western portion. As the two areas are separated by only 15 m, it is feasible that they represent different loci of site utilization; the western portion as a primary activity area and the eastern portion as a refuse area. As such, the investigation of both areas would provide for the interpretation of site use and formation.

Accordingly, recommendation is made for further investigations to concentrate upon the proxemic relationship of the eastern and western portions of the site. The scope of work for the testing phase provided for mechanical removal of the upper 40 cm of recent alluvial fill prior to manual excavation. With the recovery of a Hare biface, faunal material, and several sherds of aboriginal ceramics at 20–40 cm bs, the use of heavy machinery would not be advised for data recovery investigations at this site.

69

References Cited

Ahr, S. W., and J. T. Abbott 1999 Archeological Survey of the Proposed State Highway 36 Bridge Replacement over the Little River and Road Widening Between Travis Street in Cameron to US 77, Milam County . CSJ: 0185-04-033 & 034. Texas Department of Transportation, Austin.

Bousman, C. B. 1998 Paleoenvironmental Change in Central Texas: The Palynological Evidence. Plains Anthropologist 43(164):201–219.

Bryant, V. M., Jr. 1977 A 16,000 Year Pollen Record of Vegetation Change in Central Texas. Palynology 1:143–155.

Camper, H. A. 1991 Pollen Analysis of the Patschke Bog. Unpublished Master of Science thesis in Botany, Texas A&M University, College Station.

Chipman, D. E. 1992 Spanish Texas 1519-1821. University of Texas Press. Austin.

Collins, M. B. 1995 Forty Years of Archeology in Central Texas. Bulletin of the Texas Archeological Society 66:361–400.

Dillehay, T. D. 1974 Late Quaternary Bison Population Changes on the Southern Plains. Plains Anthropologist 19(65):180– 196.

Ellis, G. L. 1997 Amino Acid Racemization Dating of Amblema plicata from the Corn Creek Middens. In Hot Rock Cooking on the Greater Edwards Plateau: Four Burned Rock Midden Sites in West Texas , by S. L. Black, L. W. Ellis, D. G. Creel, and G. T. Goode. Appendix G, Volume 2. Studies in Archaeology 22 and Archeology Studies Program Report 2. Texas Archeological Research Laboratory and Texas Department of Transportation, Austin.

Ellis, G. L., G. A. Goodfriend, J. T. Abbott, R. E. Hare, and D. W. Von Endt 1996 Assessment of Integrity and Geochronology of Archaeological Sites Using Amino Acid Racemization in Land Snails: Examples from Central Texas. Geoarchaeology 11:189–213.

Everitt, B. S. 1977 The Analysis of Contingency Tables. John Wiley and Sons, New York.

Fickel, W., Jr. 1993 Letter Report: Archeological Survey of Streambank Erosion along the Little River, Cameron, Milam County, Texas. Ft. Worth: COE-FWD.

71 Gilmore, K. K. 1969 The San Xavier Missions: A Study in Historical Site Identification. Report Number 16. State Building Commission, Archeological Program.

Goodfriend, G. A. 1987 Evaluation of Amino-Acid Racemization/Epimerization Dating Using Radiocarbon-Dated Fossil Land Snail Shells. Geology 15:698–700. 2000 Appendix F: Rabdotus Shell A/I Ratios. In The Lino Site: A Stratified Late Archaic Campsite in a Terrace of the San Idelfonzo Creek, Webb County, Southern Texas, by J. M. Quigg. Technical Report No. 23756. TRC Mariah Associates, Austin, Texas.

Grim, R. E. 1968 Clay Mineralogy. McGraw-Hill, New York.

Holloway, R. G., L. M. Raab, and R. Stuckenrath 1987 Pollen Analysis of Late-Holocene Sediments from a Central Texas Bog. The Texas Journal of Science 39(1):71–79.

Johnson, L. 1995 Past Cultures and Climates at Jonas Terrace: 41ME29, Medina County, Texas. Report 40. Office of the State Archeologist, Austin.

Johnson, L., and G. T. Goode 1994 A New Try at Dating and Characterizing Holocene Climates, as well as Archeological Periods, on the Eastern Edwards Plateau. Bulletin of the Texas Archeological Society 65:1–51.

Kintigh, K. W. 1992 Tools for Quantitative Archaeology . Tempe, Arizona.

Mahoney, R. B., and R. G. Moore 1998 Data Recovery at 41HR817 on Clear Creek, Harris County, Texas . Report of Investigation Number 227. Moore Archeological Consulting, Inc., Houston.

Perttula, T. K. n.d. The Ceramic and Lithic Assemblage from a Late Prehistoric Site (41RT510) in the Post Oak Savanna, Robertson County, Texas. MS in preparation. 1999 Ceramic Evidence for Prehistoric Caddoan Trade and Exchange. Mini-NSF Grant Proposal on file, Research Reactor Center, University of Missouri, Columbia.

Prewitt, E. R. 1981 Cultural Chronology in Central Texas. Bulletin of the Texas Archeological Society 52:65–89. 1982 Archeological Investigations at the Loeve-Fox, Loeve and Tombstone Bluff Sites in the Granger Lake District of Central Texas . Archeological Investigations at the San Gabriel Reservoir Districts, CentraTexas, Volume 4. Institute of Applied Sciences, North Texas State University, Denton. 1985 From Circleville to Toyah: Comments on Central Texas Chronology. Bulletin of the Texas Archeological Society 54:201–238. 1995 Distributions of Typed Projectile Points in Texas. Bulletin of the Texas Archeological Society 66:83– 173.

72 Potzger, J. E., and B. C. Tharp 1954 Pollen Study of Two Bogs in Texas. Ecology 35:462–466.

Prikryl, D. J. 2001 Fiction and Fact about the Titskanwatits, or Tonkawa, of East-Central Texas. Bulletin of the Texas Archeological Society 72 (in press).

Quigg, J. M. 1999 Archaeological Monitoring at Upland Sites 41ZP39 and 41ZP176 for TransTexas Gas Well Site USA #3, Falcon Reservoir, Zapata County, Texas. Technical Report 23224. TRC Mariah and Associates, Austin, Texas.

Ramsey, C. B. 2000 OxCal Program Version 3.5. Radiocarbon Accelerator Unit, University of Oxford, Oxford, UK.

Rogers, R. 1999 Excavations at the Walleye Creek Site (41LE57), Lee County, Texas . Document Number 981670. Espey, Huston & Associates, Inc., Austin.

Rogers, R., and S. M. Kotter 1995 Archaeological Investigations at the Chesser Site (41LE59), Lee County, Texas . Document Number 950209. Espey, Huston & Associates, Inc., Austin.

Story, D., and H. J. Shafer 1965 1964 Excavations at Waco Reservoir, McLennan County, Texas: The Baylor and Britton Sites . Miscellaneous Papers No. 6. Texas Archeological Salvage Project, The University of Texas at Austin.

Stuvier, M., and P. J. Reimer 1993 Extended 14C Database and Revised CALIB Radiocarbon Calibration Program. Radiocarbon 35:215– 230.

Suhm, D., and E. B. Jelks 1962 Handbook of Texas Archeology: Type Descriptions . Texas Archeological Society Special Publication Number 1 and Texas Memorial Museum Bulletin Number 4, Austin.

Texas State Department of Highways and Public Transportation (TSDHPT) 1990 Cultural Resources Survey Milam County F.M. 485: From Tyson Creek to the Brazos River . Letter Report. TSDHPT, Austin.

Turner, E. S., and T. R. Hester 1993 A Field Guide to Stone Artifacts of Texas Indians . Gulf Publishing Company, Houston, Texas.

Turpin, J., and S. M. Carpenter 1994 Continuing Cultural Resource Management at Calvert Mine: 1993 Prehistoric Site Testing, Robertson County. Technical Series No. 42. Texas Archeological Research Laboratory, The University of Texas at Austin.

73 Turpin, S. A., L. C. Bement, and D. G. Robinson 1991 All American Pipeline Project: Final Status Report. Texas Archeological Research Laboratory, Austin.

Van Horn, D. M., J. R. Murray, and J.V. Linscheid 1993 A Method for Effectively Screening Some Clay Matrices. In Practical Archaeology, Field and Labo- ratory Techniques and Archaeological Logistics, edited by B. D. Dillon. Institute of Archaeology, University of California, Los Angeles.

Ward, G. K., and S. R. Wilson 1978 Procedures for Comparing and Combining Radiocarbon Age Determinations: A Critique. Archaeometry 20:19–31.

Wilson, S. R., and G. K. Ward 1981 Evaluation and Clustering of Radiocarbon Age Determinations: Procedures and Paradigms. Archaeometry 21:19–39.

74 Appendix A

Soil - Stratigraphic Descriptions BHT-24 (Block 1, 41MM341); elevation 85.7 to 86.1 m; floodplain; calcareous throughout; A/C-Bk2 hori- zons from middle south wall, Ab2-Bk3 horizons from north wall on west end.

A/C 0-15 cm; (Unit 3); gray (10YR 3.5/1) clay; moderate fine and medium angular blocky; very hard; few clay clasts, 0.5 cm; clear smooth. A1 15-36 cm; gray (10YR 3.5/1) clay; moderate medium angular blocky; firm; common clay clasts, 0.5 cm; gradual smooth. A2 36-87 cm; dark gray (10YR 3/1) clay; few brown (10YR 5/3) pockets; moderate coarse angular blocky; firm; few clay clasts, 0.5 cm; gradual smooth. Bk1b1 87-113 cm; (Unit 2); dark grayish brown (10YR 4/2) clay; few brown (10YR 5/3) pockets; weak coarse prismatic; firm; 3% calcium carbonate filaments; gradual smooth. Bk2 113-137 cm; dark grayish brown (10YR 4/2) clay; weak coarse prismatic; firm; 3% calcium car- bonate filaments, 1% carbonate nodules, 0.5 cm; clear smooth. Ab2 137-177 cm; (Unit 1); dark gray (10YR 3/1) clay; few dark grayish brown (10YR 3.5/2) pockets; few fine faint yellowish brown (10YR 3/4) soft iron masses; weak coarse subangular blocky; firm; 1% calcium carbonate filaments; 1% calcium carbonate nodules, 0.5 cm; gradual smooth. Bk1 177-219 cm; dark grayish brown (10YR 4/2) clay; common fine very dark grayish brown (10YR 3/2) pockets; weak coarse angular blocky; firm; few faint slickensides; 2% calcium carbonate nodules, 0.5 to 1 cm; 1% calcium carbonate filaments; gradual smooth. Bk2 219-258 cm; dark grayish brown (10YR 4/2) clay; few very dark grayish brown (10YR 3/2) pock- ets; weak coarse prismatic; very firm; 3% calcium carbonate filaments; 2% calcium carbonate nodules, 0.5 to 1 cm; gradual smooth. Bk3 258-284 cm; light olive brown (2.5Y 5/3) clay; 20% fine faint gray (10YR 5/1) iron depletions; few fine distinct yellowish brown (10YR 5/4) soft iron masses; weak coarse prismatic; very firm; 2% calcium carbonate filaments; 2% calcium carbonate nodules, 0.5 to 1 cm. BHT-15 (Block 2, 41MM341); elevation 86.1 to 86.2 m; floodplain; calcareous throughout; A/C-Bk2 hori- zons from middle east wall, Akb2-Bk3 horizons from east wall below test unit.

A/C 0-24 cm; (Unit 3); dark gray (10YR 3.5/1) clay; moderate medium angular blocky and platy; very hard; clear smooth. A 24-53 cm; dark gray (10YR 3.5/1) clay; few brown (10YR 5/3) pockets; weak coarse angular blocky; firm; common clay clasts, 0.5 cm; clear smooth. Ab1 53-79 cm; (Unit 2); very dark gray (10YR 3/1) clay; few brown (10YR 5/3) pockets; moderate medium and coarse angular blocky; firm; few clay clasts, 0.5 cm; gradual smooth. Bk1 79-113 cm; dark grayish brown (10YR 3.5/2) clay; 10% brown (10YR 5/3) pockets; weak coarse prismatic; hard; 2% calcium carbonate filaments; gradual smooth. Bk2 113-140 cm; dark grayish brown (10YR 4/2) clay; weak medium and coarse prismatic; firm; 3% calcium carbonate filaments; clear smooth Akb2 140-159 cm; (Unit 1); dark gray (10YR 3/1) clay; 10% dark grayish brown (10YR 4/2) pockets; weak coarse angular blocky; very firm; gradual smooth.

76 Bk1 159-188 cm; dark gray (10YR 3.5/1) clay; 5% fine distinct gray (10YR 5/1) iron depletions along ped faces; few fine distinct dark yellowish brown (10YR 3/4) soft iron masses; moderate coarse prismatic; very firm; 2% calcium carbonate filaments; 2% calcium carbonate nodules, 0.5 to 1 cm; gradual smooth. Bk2 188-223 cm; dark grayish brown (10YR 3.5/2) clay; 5% fine distinct gray (10YR 5/1) iron deple- tions along ped faces; moderate coarse prismatic; very firm; 3% calcium carbonate nodules; 1% calcium carbonate filaments; gradual smooth. Bk3 223-280 cm; light olive brown (2.5Y 5/3) clay; few fine distinct dark yellowish brown (10YR 4/4) soft iron masses; 5% fine distinct gray (10YR 5/1) iron depletions along ped faces; weak coarse prismatic; extremely firm; 3% calcium carbonate nodules; 1% calcium carbonate filaments.

BHT-5; elevation ÷85.5 m; slough; calcareous throughout; east wall.

A/C 0-18 cm; (Unit 3); light olive brown (2.5Y 5/3), grayish brown (2.5Y 5/2), and very dark gray (2.5Y 3/1) clay; massive; extremely firm; few siliceous pebbles, 0.3 to 0.5 cm; abrupt smooth. A 18-47 cm; dark gray (10YR 3.5/1) clay loam; weak coarse angular blocky; firm; common clay clasts, 0.5 cm; few siliceous pebbles, 0.3 to 0.5 cm; gradual smooth. Bw 47-96 cm; dark gray (10YR 3.5/1) clay loam; weak coarse angular blocky; firm; two faint yellow- ish brown (10YR 5/4) beds, 1 cm; common clay clasts, 0.5 cm; clear smooth. Ab 96-120 cm; (Unit 2); very dark gray (10YR 3/1) clay; few yellowish brown (10YR 5/4) pockets; weak coarse angular blocky; very hard; common clay clasts, 0.5 cm; gradual smooth. Bw 120-156 cm; dark gray (10YR 3.5/1) clay; few yellowish brown (10YR 5/4) pockets; weak coarse angular blocky; very hard; common clay clasts, 0.5 cm; gradual smooth. C 156-240 cm; bedded clays and loams; clays-very dark and dark gray (10YR 3/1, 4/1), many clay clasts, 3 to 10 cm; loams-light brownish gray (10YR 6/2) and light gray (10YR 7/2) very fine sandy loam and silt loam, 0.2 to 0.5 cm. BHT-16; elevation 85.4 m; slough; calcareous throughout; east wall.

A/C1 0-9 cm; (Unit 3); very dark gray (10YR 3/1) clay; distinct bedding planes; hard; abrupt smooth. A/C2 9-38 cm; dark gray (10YR 3.5/1) clay; massive; plastic; 5% siliceous pebbles, 0.5 to 2 cm; abrupt wavy. Ab1 38-87 cm; dark grayish brown (10YR 3.5/2) clay; faint brown (10YR 5/3) bed in middle; weak coarse angular blocky; very firm; common clay clasts, 0.5 cm; gradual smooth. Bw 87-109 cm; dark grayish brown (10YR 4/2) clay; weak coarse angular blocky; very firm; common clay clasts, 0.5 cm; gradual smooth. Bss 109-131 cm; dark grayish brown (10YR 4/2) clay; moderate medium angular blocky; very firm; common distinct slickensides; abrupt smooth. A1b1 131-165 cm; (Unit 2); dark grayish brown (2.5Y 4/2) clay; 30% brown (10YR 5/3) pockets; light brownish gray (10YR 6/2) silt loam lamination at top and bottom, 3 mm; faint light brownish gray (10YR 6/2) silt loam bed in middle, 2 cm; weak coarse angular blocky; very firm; few distinct slickensides; clear smooth.

77 A2 165-210 cm; dark gray (10YR 3.5/1) clay; few brown (10YR 5/3) pockets; weak coarse angular blocky; extremely firm; few distinct slickensides; 1% calcium carbonate nodules, 0.5 cm; gradual smooth. Bk 210-245 cm; grayish brown (2.5Y 4.5/2) clay; weak coarse angular blocky; extremely firm; 5% fine faint dark gray (2.5Y 4/1) iron depletions along ped faces; few fine distinct dark yellowish brown (10YR 4/4) soft iron masses. BHT-18; elevation 85.5 m; slough; calcareous throughout; east wall.

A/C 0-17 cm; (Unit 3); dark grayish brown (2.5Y 4/2) clay; very dark gray (2.5Y 3/1) bed in middle, 2 cm; weak coarse angular blocky; very firm; clear smooth. A 17-53 cm; dark grayish brown (10YR 4/2) clay; moderate coarse angular blocky; firm; common clay clasts, 0.5 cm; clear smooth. Ab1 53-72 cm; dark grayish brown (10YR 3.5/2) clay; weak coarse angular blocky; firm; common clay clasts, 0.5 cm; gradual smooth. Bw 72-120 cm; dark grayish brown (2.5Y 4/2) clay; 3% brown (10YR 5/3) pockets; weak coarse pris- matic; very firm; few faint slickensides; few clay clasts, 0.5 cm; clear smooth. C 120-143 cm; dark grayish brown (10YR 3.5/2) clay; 5% brown (10YR 5/3) pockets; weak coarse prismatic; very hard; few clay clasts, 0.5 cm; gradual smooth. Ab2 143-180 cm; (Unit 2); very dark grayish brown (10YR 3/2) clay; 3% dark grayish brown (10YR 4/ 2) pockets; weak coarse angular blocky; very hard; gradual smooth. Bk180-211 cm; dark grayish brown (2.5Y 4/2) and light olive brown (2.5Y 5/3) clay; few fine faint dark yellowish brown (10YR 4/4) soft iron masses; weak coarse prismatic; very hard; 1% calcium carbonate nodules, 0.5 cm; abrupt smooth. Bg 211-235 cm; very dark gray (10YR 3/1) clay; 15% brown (10YR 4.5/3) and dark grayish brown (10YR 4/2) pockets; few fine faint gray (2.5Y 5/1) iron depletions; weak coarse angular blocky; extremely firm; many clay clasts, 0.5 cm; gradual smooth. BC 235-288 cm; dark gray (10YR 4/1) clay; weak coarse angular blocky to massive; extremely firm. BHT-7; elevation 85.5 m; slough; calcareous throughout; east wall.

A1 0-33 cm; (Unit 3); dark gray (2.5Y 3.5/1) clay; weak medium and coarse subangular blocky; firm; few medium distinct dark yellowish brown (10YR 4/4) soft iron masses; 5% fine faint gray (2.5Y 5/ 1) iron depletions along ped faces; clear smooth. A2 33-51 cm; very dark gray (2.5Y 3/1) clay; moderate medium and coarse angular blocky; very firm; common medium distinct dark yellowish brown (10YR 4/4) soft iron masses; few faint gray (2.6Y 6/1) laminations; gradual smooth. Bw 51-64 cm; dark gray (2.5Y 4/1) and light brownish gray (2.5Y 6/2) clay loam; weak coarse angular blocky; very hard; clear smooth. Ab1 64-85 cm; very dark gray (10YR 3/1) clay; few light brownish gray (2.5Y 6/2) pockets; weak coarse angular blocky; hard; few siliceous pebbles, 0.2 to 0.5 cm; gradual smooth. Bw 85-104 cm; dark grayish brown (2.5Y 4/2) clay loam; 10% light brownish gray (2.5Y 6/2) pockets; weak coarse angular blocky; hard; few siliceous pebbles, 0.2 to 0.5 cm; gradual smooth.

78 C 104-114 cm; dark gray (10YR 3.5/1) clay loam; 2% faint light brownish gray (2.5 6/2) laminations; hard; clear smooth. Ab2 114-133 cm; ; (Unit 2); dark gray (10YR 3.5/1) clay; 10% light brownish gray (2.5Y 6/2) pockets; weak coarse angular blocky; hard; few fine distinct dark brown (7.5YR 3/4) soft iron masses; gradual smooth. C 133-141 cm; dark gray (10YR 4/1) clay; common medium distinct dark yellowish brown (10YR 3/ 4) soft iron masses; hard; clear smooth. Ab3 141-166 cm; very dark gray (10YR 3/1) clay; weak coarse angular blocky; very firm; common medium distinct dark yellowish brown (10YR 4/4) soft iron masses; gradual smooth. Bg 166-218 cm; dark gray (10YR 4/1) clay; common medium distinct dark yellowish brown (10YR 4/ 6) soft iron masses; 5% faint light brownish gray (2.5Y 6.2) laminations; weak coarse angular blocky; firm; gradual smooth. Cg1 218-247 cm; laminated very dark gray and dark gray (2.5Y 3/1, 4/1) clay; common medium distinct dark yellowish brown (10YR 4/4) soft iron masses; firm; clear smooth. Cg2 247-280 cm; dark gray (2.5Y 4/1) clay; massive; very firm; many medium distinct dark yellowish brown (10YR 4/4) soft iron masses.

BHT-8; elevation 85.7 to 85.9 m; floodplain; calcareous throughout; south wall.

A/C 0-22 cm; (Unit 3); massive gray (10YR 5/1) with faint yellowish brown (10YR 5/4) laminations; friable; abrupt wavy. Ab1 22-39 cm; (Unit 2); very dark gray (10YR 3.5/1) clay; weak coarse angular blocky; very firm; gradual smooth. Bw 39-75 cm; dark gray (10YR 4/1) clay; weak coarse angular blocky; extremely firm; few medium faint brown (10YR 4/3) soft iron masses; clear smooth. Ab2 75-88 cm; (Unit 1); very dark gray (2.5Y 3/1) clay; few light olive brown (2.5Y 5/3) pockets; weak coarse angular blocky; extremely firm; gradual smooth. Bk1 88-119 cm; brown (10YR 4/3) clay; 5% very dark gray (10YR 3/1) pockets; weak coarse prismatic; very hard; 2% calcium carbonate filaments; gradual smooth. Bk2 119-143 cm; dark grayish brown (10YR 4/2) clay; 5% very dark gray (10YR 3/1) pockets; weak coarse prismatic; very hard; 5% calcium carbonate filaments; gradual smooth. Bk3 143-200 cm; gray (10YR 4/1) clay; weak medium prismatic; very hard; moderate medium distinct dark yellowish brown (10YR 4/4/) soft iron masses; 5% calcium carbonate filaments; gradual smooth. Bk4 200-281 cm; gray (10YR 4/1) clay; weak medium prismatic; very hard; moderate medium distinct dark yellowish brown (10YR 4/4) soft iron masses; 2% calcium carbonate filaments; gradual smooth. Bk5 281-300 cm; light olive brown (2.5Y 5/4) clay; many medium distinct dark gray (2.5Y 4/1) pockets; weak coarse prismatic; very hard; 3% calcium carbonate filaments. BHT-13; elevation 85.9 m; floodplain; calcareous throughout; east wall.

A/C 0-14 cm; (Unit 3); backhoe trench backfill, variegated colors; massive; gradual smooth.

79 Ab1 14-29 cm; dark grayish brown (10YR 3.5/2) clay; weak medium angular blocky; firm; gradual smooth. Bw1 29-62 cm; dark grayish brown (10YR 4/2) clay; weak medium and coarse angular blocky; firm; gradual smooth. Bw2 62-117 cm; grayish brown (2.5Y 5/2) clay; weak medium and coarse angular blocky; firm; gradual smooth. Ab2 117-161 cm; (Unit 2); dark grayish brown (2.5Y 4/2) clay; weak coarse prismatic to moderate medium angular blocky; firm; few faint slickensides; gradual smooth. Bk1 161-191 cm; dark grayish brown (10YR 4/2) clay; moderate medium prismatic; very firm; 2% calcium carbonate filaments; gradual smooth. Bk2 191-230; grayish brown (10YR 4.5/2) clay; weak coarse prismatic; very firm; 2% calcium carbon- ate nodules, 0.5 to 1 cm. BHT-12; elevation 85.6 m; floodplain; calcareous throughout; east wall.

A/C 0-13 cm; (Unit 3); dark grayish brown (10YR 3.5/2) clay; weak medium subangular blocky; very firm; clear smooth. Ab1 13-35 cm; very dark grayish brown (10YR 3/2) clay; weak medium subangular blocky; very firm; gradual smooth. C 35-61 cm; grayish brown (10YR 4/2.5) clay; massive; firm; faint beds; clear smooth. Ab2 61-95 cm; (Unit 2); very dark grayish brown (10YR 3/2) clay; weak coarse angular blocky; very hard; 5% light yellowish brown (10YR 5/4) pockets; gradual smooth. Bk1 95-141 cm; grayish brown (10YR 4.5/2) clay; weak coarse angular blocky; very hard; 5% light yellowish brown (10YR 5/4) pockets; 1% calcium carbonate filaments; 1% calcium carbonate fila- ments; gradual smooth. Bk2 141-200 cm; gray (10YR 5/1) clay; weak coarse angular blocky; very hard; 1% calcium carbonate filaments; 1% calcium carbonate filaments; few faint slickensides; gradual smooth. BHT-2; elevation 85.5-85.6 m; slough; calcareous throughout; south wall.

A1 0-19 cm; (Unit 3); very dark grayish brown (2.5Y 3/2) clay; strong medium subangular; firm; com- mon siliceous pebbles, 0.5 cm; gradual smooth. A2 19-36 cm; dark grayish brown (2.5Y 3/2) clay; moderate coarse subangular; firm; few siliceous pebbles, 0.5 cm; few clay clasts, 0.5 cm; gradual smooth. Bw1 36-75 cm; dark grayish brown (2.5Y 4/2) clay; weak coarse subangular blocky; firm; few siliceous pebbles, 0.5 cm; faint light brownish yellow (2.5Y 6/4) laminations in lower 5 cm; abrupt smooth. Bw2 75-92 cm; dark grayish brown (2.5Y 3.5/2) clay; weak medium angular blocky; very firm; 5% light yellowish brown (2.5Y 6/3) pockets; clear smooth. Bw/C 92-129 cm; dark grayish brown (2.5Y 4/2) clay loam; common distinct pale brown (10YR 6/3) laminations; many clay clasts, 0.5 to 1.5 cm; abrupt smooth. C1 129-143 cm; faintly laminated pale brown (10YR 6/3) and very dark grayish brown (10YR 3/2) silty clay loam; hard; abrupt smooth.

80 C2 143-153 cm; few faint pale brown (10YR 6/3) and very dark grayish brown (10YR 3/2) laminated clays; very hard; many clay clasts, 0.5 to 2 cm; few fine distinct brown (7.5YR 4/4) soft iron masses; abrupt smooth. C3 153-202 cm; laminated pale brown (10YR 6/3) and dark grayish brown (10YR 4/2) silt loam; hard; 5% very dark grayish brown (10YR 3/2) pockets; abrupt smooth. C4 202-217 cm; few distinct pale brown (10YR 6/3) and very dark grayish brown (10YR 3/2) lami- nated silt loams; slightly hard.

BHT-11; elevation 85.3 m; north slough; calcareous throughout; north wall.

fill 0-33 cm; (Unit 3); very dark gray (10YR 3.5/1) clay; weak coarse angular blocky; very firm; com- mon siliceous pebbles, 0.5 to 2 cm; abrupt wavy. fill 33-72 cm; very dark gray and black (2.5Y 3/1, 2/1) and brownish yellow (10YR 6/6) clay; massive; firm; many siliceous pebbles, 0.5 to 2 cm; 2% detrital calcium carbonate nodules, 1 cm; abrupt wavy. Bw 72-99 cm; dark gray (2.5Y 3.5/1) clay; common fine distinct brown (7.5YR 4/4) soft iron masses; weak coarse angular blocky; firm; few clay clasts, 0.5 cm; clear smooth. Assgb1 99-132 cm; (Unit 2); very dark gray (2.5Y 3/1) clay; few fine distinct brown (7.5YR 4/4) soft iron masses; weak coarse angular blocky; very firm; common distinct slickensides; gradual smooth. Bg1 132-161 cm; gray (2.5Y 5/1) clay; common fine distinct dark yellowish brown (10YR 4/4); weak coarse angular blocky; firm; clear smooth. Bg2 161-186 cm; dark gray (2.5Y 4/1) clay; weak coarse angular blocky; very firm; common fine dis- tinct yellowish brown (10YR 5/4) soft iron masses; gradual smooth. Bg3 186-208 cm; gray (2.5Y 5/1) clay; moderate medium distinct brown (7.5YR 5/4) soft iron masses; weak coarse angular blocky; very firm; 1% calcium carbonate nodules; gradual smooth. Bg4 208-230 cm; gray (2.5Y 5/1) clay; few fine distinct yellowish brown (10YR 4/4) soft iron masses; weak coarse angular blocky; very firm. BHT-26 (Block 1, 41MM340); elevation 85.5 to 85.9 m; floodplain; calcareous throughout; south wall.

A/C 0-18 cm; (Unit 3); dark gray (10YR 3/1.5) clay; moderate coarse platy; very hard; common sili- ceous pebbles, 0.5 cm; clear smooth. Ab1 18-43 cm; (Unit 2); dark gray (10YR 3.5/1) clay; weak coarse angular blocky; very hard; gradual smooth. Bw 43-71 cm; dark grayish brown (2.5Y 3.5/2) clay; weak coarse angular blocky; firm; few siliceous pebbles, 0.5 cm; clear smooth. Ab2 71-94 cm; (Unit 1); dark gray (2.5Y 3.5/1) clay; weak coarse angular blocky; very firm; 3% light olive brown (2.5Y 5/3) pockets; 1% calcium carbonate filaments; gradual smooth. Bk1 94-115 cm; grayish brown (2.5Y 3.5/2) clay; 3% light olive brown (2.5Y 5/3) pockets and 1% very dark gray (2.5Y 3/1) pockets; weak coarse angular blocky; firm; 2% calcium carbonate filaments; gradual smooth.

81 Bk2 115-140 cm; gray (2.5Y 4/1) clay; weak coarse angular blocky; very firm; 1% very dark gray (2.5Y 3/1) pockets; moderate medium distinct brown (10YR 4/3) soft iron masses; 4% calcium carbonate filaments; gradual smooth. Bk3 140-177 cm; gray (2.5Y 4/1) clay; common medium distinct dark yellowish brown (10YR 4/4) soft iron masses; weak coarse angular blocky; firm; 3% calcium carbonate nodules, 0.5 to 1 cm.

BHT-B (Block 2, 41MM340); elevation 85.9 m; floodplain; calcareous throughout; late Holocene over early Holocene alluvium; east wall.

A1 0-8 cm; (Unit 3); dark gray (2.5Y 3.5/1) clay; weak coarse angular blocky; very firm; clear smooth. A2 8-31 cm; very dark gray (2.5Y 3.5/1) clay; weak coarse angular blocky; very firm; clear wavy. Bw1b1 31-60 cm; (Unit 2); very dark gray (10YR 3/1) clay; weak coarse angular blocky; very firm; clear wavy. Bw2 60-74 cm; very dark gray (10YR 3/1) clay; weak coarse angular blocky; very firm; clear smooth. Akb2 74-90 cm; (Unit 1); black (2.5Y 2.5/1) clay; moderate medium prismatic to moderate medium angu- lar blocky; very hard; 2% dark grayish brown pockets; 2% calcium carbonate filaments; gradual smooth. Bk1 90-116 cm; very dark gray (10YR 3/1) clay; moderate medium prismatic to moderate medium angu- lar blocky; very hard; 8% calcium carbonate filaments; gradual smooth. Bk2 116-139 cm; dark gray (2.5Y 3.5/1) clay; few black (2.5Y 2.5/1) pockets; 30% brown (10YR 5/3) pockets; moderate medium prismatic to moderate medium angular blocky; very hard; 10% calcium carbonate filaments; gradual smooth. Bk3 139-157 cm; very dark gray (2.5Y 3/1) clay; 30% brown (10YR 5/3) pockets; moderate medium prismatic to moderate medium angular blocky; very hard; 10% calcium carbonate filaments; gradual smooth. Bk4 157-180 cm; yellowish brown (10YR 5/4) clay; few fine distinct dark yellowish brown (10YR 4/6) soft iron masses; common medium distinct dark gray (2.5Y 4/1) iron depletions along ped faces; weak coarse prismatic; very hard; 10% calcium carbonate filaments. BHT-A (Block 3, 41MM340); elevation 85.4 to 85.7 m; floodplain; calcareous throughout; north wall; (missing upper 30 cm).

Bw 0-31 cm; (Unit 3); dark gray (10YR 3/1.5) clay; weak coarse angular blocky; very firm; clear smooth. Ab1 31-58 cm; (Unit 1); very dark gray (10YR 3/1) clay; moderate medium prismatic to moderate me- dium angular blocky; very hard; 1% dark grayish brown (10YR 4/2) pockets; 1% calcium carbonate filaments; gradual smooth. Akb2 58-79 cm; very dark gray 10YR 3/1) clay; 2% dark grayish brown (10YR 4/2) pockets; weak coarse prismatic; very hard; 5% calcium carbonate filaments; gradual smooth. Bk2 79-117 cm; dark gray (10YR 4/1) clay; 5% dark yellowish brown (10YR 4/4) pockets; weak coarse prismatic; very hard; 5% calcium carbonate filaments; gradual smooth.

82 Bk3 117-134 cm; dark gray (10YR 4/1) clay; weak coarse prismatic; very hard; moderate medium dis- tinct yellowish brown (10YR 5/4) soft iron masses; 5% calcium carbonate filaments.

CB-6; modern levee; calcareous throughout.

C 0-2 cm; (Unit 3); pale brown (10YR 6/3) fine sandy loam; single grain; abrupt wavy. A 2-29 cm; very dark gray (10YR 3/1) clay; moderate fine and medium subangular blocky; very hard; clear smooth. Bw1 29-49 cm; dark gray (10YR 3.5/1) clay; moderate medium and coarse subangular blocky; very hard; faint bed in lower part, 2 cm; clear smooth. Bw2 49-71 cm; dark gray (10YR 3.5/1) clay; 30% brown (10YR 5/3) pockets; moderate medium and coarse subangular blocky; hard; few faint laminations; brown clay bed at base, 2 cm; clear smooth. Bw3 71-87 cm; dark gray (10YR 3.5/1) and brown (10YR 5/3) laminated clay loam and loam; weak coarse angular blocky; hard; common faint laminations; abrupt smooth. C1 87-123 cm; brown (10YR 5/3) loam; 30% dark gray (10YR 4/1) pockets; massive; hard; common faint laminations; abrupt smooth. C2 123-140 cm; dark gray (10YR 4/1) clay loam; weak coarse prismatic; very hard; 30% brown (10YR 5/3) pockets; common fine distinct strong brown (7.5YR 4/6) iron pore linings; clear smooth. C3 140-159 cm; brown (10YR 5/3) loam; 20% dark gray (10YR 4/1) pockets; massive; hard; few faint laminations; 2 cm clay bed at base with common fine distinct strong brown (7.5YR 4/6) iron pore linings; abrupt smooth. C4 159-194 cm; dark gray (10YR 4/1.5) clay loam; 10% brown (10YR 5/3) pockets; weak coarse prismatic; very hard; common fine distinct strong brown (7.5YR 4/6) iron pore linings; clear smooth. C5 194-220+ cm; pale brown (10YR 6/3) silt loam; 20% dark gray (10YR 4/1) pockets; few distinct beds and laminations of clay and silt loam; massive; hard; common fine distinct strong brown (7.5YR 4/6) iron pore linings.

83

Appendix B

Radiocarbon Results 86 87 88 Appendix C

Vertebrate Faunal Remains

Barbara A. Meissner Appendix C: Vertebrate Faunal Remains Barbara A. Meissner

This report is intended to provide an assessment of the identifiable vertebrate fauna recovered from 41MM340 and 41MM341 during the project. Only the most readily identifiable bone was closely examined.

In general, the bone is in moderate to good condition. There is little evidence of extensive atmospheric or chemical weathering. Most of the bone is in very small pieces, but the general impression is that much of this breakage was done while the bone was fresh. This will be assessed in the final report. It appears, however, that the small number of bone recovered is not due to poor preservation. It seems likely that there were very few bones left in these sites, and that most of this bone was very heavily processed.

Seven genera of mammals were identified (see Table C-1). Deer, cottontail and jackrabbit were the most com- mon. With two exceptions, all identified mammals are wild animals commonly found in the area today. One possible exception is Canis, which could be dog, coyote, or wolf. All of the Canis bone was about the size of a coyote, but could be large dog or small wolf.

The other exception was a single molar belonging to either a goat (Capra) or a sheep (Ovis). Although there are wild members of these genera in Texas, they are all adapted to mountainous environments and are only found in far West Texas. This tooth is probably from a domestic animal. Its location in Level 6 (50–60 cm below datum) presents a small problem. The earliest domestic sheep and goat in the area would have been in association with Mission San Xavier. It seems most likely that the goat/sheep tooth was from a much more recent animal. It could have fallen into a deep crack or been tracked into the unit in mud on someone’s boot.

Bird bone was not uncommon in this collection, but most of it was highly fragmented, and none could be identified. Most of it appeared to be of duck size or larger. One bird bone bead, made on a long bone of a large bird (at least goose-sized, and possibly as large as a hawk) was identified.

Fragments of turtle carapace, many of which were burned, were common. Twenty-nine fragments were recov- ered at 41MM340, of which 16 were from a single Pseudomys individual and the rest were unidentifiable. Sixteen carapace fragments were identified at 41MM341. These included Pseudomys, Terrepene, and Trionyx genera, as well as a number of unidentified fragments.

The only fish identified was catfish (Ictalurus sp.). Most of this was the bones of a single medium-sized head, excavated in Block 1 of 41MM340, between 10 and 20 cm below datum. A few fragments of unidentified fish were also recovered, but these were rare. All were in 41MM340.

In conclusion, very little bone was recovered during this project. The condition of the bone that was found suggests that poor preservation is not the reason for the small numbers of bone. Much of the bone is in very small pieces and was probably heavily processed. The majority of the unidentified bone appears to be from animals of the deer/antelope size category, however at least two unidentified bones could have come from a bison-sized animal, based on cortical thickness. Most of the bone is from wild mammals, from genera still found in the area today, however, there are numerous bird bones and turtle shell fragments as well.

90 Table C-1. Taxa identified from 41MM340 and 41MM341

Genus Common Name 41MM340 41MM341 Mammals Dog, Coyote, Canis Wolf X Capra/Ovis Goat or sheep X Didelphis Opossum X Lepus Jackrabbit X X Odocoileus Deer X Procyon Raccoon X Sylvilagus Cottontail X X

Reptiles Pseudomys Pond Slider X X Terrepene Box Turtle X Soft-shelled Trionyx Turtle X

Fish Ictalurus Catfish X

91

Appendix D

Freshwater Mussel Shell

Robert G. Howells Appendix D: Freshwater Mussel Shell Robert G. Howells

Introduction Freshwater mussels (Family Unionidae) have been used by Native Americans for food, tools, and decoration (Howells et al. 1996; Neck 1982) and are frequently associated with archeological sites. Following is a prelimi- nary analysis of unionid shell remains obtained by the Center for Archaeological Research (CAR), University of Texas at San Antonio, from sites 41MM340 and 41MM341 near Little River (a Brazos River tributary), Milam County, Texas, in early 2000.

Materials and Methods Specimens were obtained by CAR during archeological excavations at these two adjacent sites. Four lots of selected shells (49–50/lot) were selected for preliminary analysis. Specimens were identified using Howells et al. (1996) and by comparison to shells of species collected by the author in or near the Central Brazos River Drainage. Taxonomic terminology follows Howells et al. (1996). Additionally, annual statewide field-survey reports produced by Texas Parks and Wildlife Department (Howells 1995, 1996, 1997) were reviewed for com- parative mussel assemblage composition by species and number. However, comparative data were selected only from river and stream sites with the exclusion of reservoir collections (no lakes or reservoirs were present in this area in prehistoric times).

Results and Discussion A total of seven species of freshwater mussels were found among the 198 shells and valves examined (Table D- 1). These included threeridge (Amblema plicata), smooth pimpleback (Quadrula houstonensis), southern mapleleaf (Quadrula apiculata), Tampico pearlymussel (Cyrtonaias tampicoensis), bleufer (Potamilus purpuratus), false spike (Quincuncina mitchelli), and pistolgrip (Tritogonia verrucosa). Among these, threeridge and smooth pimpleback were far more abundant than other taxa. All of the identified species are (or were) typical of Brazos River unionid fauna (Howells et al. 1996; Strecker 1931).

Shells and valves examined were generally in very good condition. Some displayed signs of erosion since the time of the animal’s death, but most were well preserved. Most had small to large portions of their disks absent, but with no clear evidence of human manipulation, including heating. Most specimens were smaller adults, but with a few juveniles or much larger adults occasionally represented.

Species composition varied between depth and site. Threeridge species decreased in abundance from 84 percent of the total at 43–53 cm below the surface (cm bs) at site 41MM340 to 66 percent at 94–104 cm bs and 44.9 percent at 124–134 cm bs, and was only 43.7 percent of the total collection from site 41MM341. Conversely, smooth pimpleback increased from 12 percent at 43–53 cm bs to 38.8 percent of the total at 124–134 cm bs at site 41MM340, and was 59.2 percent of the total collection from site 41MM341. Southern mapleleaf was present in all four lots, but only included 1–6 specimens/lot. Other taxa were even less frequent.

94 Table D-1. Identification of selected freshwater mussel shell remains recovered from archeological sites 41MM340 and 41MM341

Ele vation Percent of Sample # Site Block Unit (cmbs) Lot# Zone Common Name Scientific Name Shells Valves Total Total 1 41MM340 2 C 43-53 78 2 Threeridge Amblema plicata 1414284 Smooth pimpleback Quadrula houstonensis 156 12 Southern mapleleaf Quadrula apiculata 011 2 Tampico pearlymussel Cyrtonaias tampicoensis 011 2 2 41MM340 3 B 94-104 181 6 Threeridge Amblema plicata 0333366 Smooth pimpleback Quadrula houstonensis 0151530 Southern mapleleaf Quadrula apiculata 011 2 Bleufer Potamilus purpuratus 011 2 3 41MM340 3 B 124-134 235 8 Threeridge Amblema plicata 0222244.9 Smooth pimpleback Quadrula houstonensis 0191938.8 Southern mapleleaf Quadrula apiculata 06612.4 False spike Quincuncina mitchelli 011 2 Pistolgrip Tritogonia verrucosa 011 2 4 41MM341 1 B 56-60 33 2 Threeridge Amblema plicata 0171734.7 Smooth pimpleback Quadrula houstonensis 0292959.2 Southern mapleleaf Quadrula apiculata 0224.1 Tampico pearlymussel Cyrtonaias tampicoensis 011 2

Combined species assemblages from a number of locations in the Central Brazos River Drainage (Table D-2) indicate that threeridge is still the most abundant unionid taxon, but is reduced from percentages found at these archeological sites with percent composition falling from 34–84 percent historically to a mean of 27 percent in recent collections. Smooth pimpleback also fell from 12–59 percent in archeological samples to less than 10 percent in recent surveys. Conversely, southern mapleleaf increased in relative abundance from 2–4 percent to nearly 25 percent.

Freshwater mussels are sensitive barometers of environmental quality and are subsequently the most-rapidly declining faunal group in North America (Howells et al. 1996; Neves 1993; Williams et al. 1993). Freshwater mussel populations in Central Texas in particular have experienced dramatic changes in distribution, abun- dance, and composition since the late 1970s (Howells et al. 1997). Endemic smooth pimpleback has been elimi- nated from many parts of its range in Central Texas in recent years (Howells et al. 1997). Data from these archeological samples suggest its decline may have started long before human modification of aquatic ecosys- tems was initiated. This apparent early decline might be associated with a general trend toward a more arid environment in the region. False spike, another endemic, has apparently not been found alive since major flood- ing and associated scouring and habitat modification in 1978. The single specimen from the archeological samples, paucity of historical data, and complete absence of living or recently-dead specimens in current studies confounds useful conclusions about it. Pistolgrip ranges as far west as the Guadalupe River and the upper Brazos Drainage, but has declined in most areas west of the Neches River; often, for reasons that are not clearly apparent.

The increase in relative abundance of southern mapleleaf between the archeological collections and the present is particularly interesting. Archeological specimens must have been collected from riverine environments, be- cause there were no lakes or reservoirs present in the area at the time of initial harvest and the species is not especially tolerant of ponds. Southern mapleleaf adapts well to many reservoirs despite having evolved in a flowing water habitat, and it can be disproportionately abundant in some reservoirs. The increase in apparent abundance seen here may reflect some other environmental shift. Again, as the region has become more arid and

95 e T ______Department 1993-1996 at selected locations in the Central Brazos River Basin U.S. 281 SH 36 S.H. 340 near Little R. near Calvert River C.R. 303 S.H. 437 U.S. 190 S.H. 50 Table D- 2. Freshwater mussel assemblages (number and percent composition) documented during surveys by Texas Parks and Wildlif Parks and Texas D- 2. Freshwater mussel assemblages (number and percent composition) documented during surveys by Table LocationCounty dateCollection Source Leon River Oct 1993 Leon HamiltonRiver Oct 1993 1995 Howells Brazos River Hamilton 1995 Howells Brazos River Aug 1994 1996 Howells 1996 McLennan Howells SandyCreek Aug 1994 1996 Howells Little Brazos Milam 1996 Howells Leon 1997River Howells Jun 1994 1997 Howells Little River 1997 Robertson Howells 1994 June 1997 Howells Little River Robertson 1996 Jan Creek Yegua Hamilton 1996 Jul Milam 1996 Jul Milam1996 Sep Washington Amblema plicata confragosus Arcidens Cyrtonaias tampicoensis Lampsilis teres fragilis Leptodea Megalonaias nervosa purpuratus Potamilus grandis Pyganodon Quadrula apiculata Quadrula houstonensis Truncilla macrodon Tritogonia verrucosa ashborad ------4 8.0-- 1 5.9 110.0-- 6 110.0-- 5.9 1 8.0------4 ashborad U.S. Highways (U.S.), state highways (S.H.), and county roads (C.R.) are indicated under location. under indicated are (C.R.) roads county and (S.H.), highways state (U.S.), Highways U.S. iant floater ------1 4.0------1 ------1 floater iant ock-pocketbook ------1 1.9 1 1.9 ------1 ock-pocketbook leufer 2 10.0 4 32.0 1 10.0 1 11.1 - - 1 2.0 1 6.3 - - 1 10.0 8 14.8 19 ellow sandshell 1 5.0-- 110.0 333.3 1 4.0------1729.623 1 333.3 110.0 5.0-- 1 sandshell ellow hreeridge 6 30.0 1 4.0 - - 2 22.2 6 24.0 13 26.0 7 43.8 13 76.5 6 60.0 7 13.0 61 ragile papershell 210.0 432.0-- 111.1---- 212.5---- 213.711 6.3----1018.555 1 212.5---- 111.1---- 936.02958.0 432.0-- 330.0-- 210.0 -- 324.0 papershell pecies mapleleaf No. %ragile No. % No. %outhern No.mooth pimpleback % No.istolgrip % 4 No. 20.0 % - No. - 5 % - 25.0 No. 1 - % 8.0 No. 1 - % 11.1 - No. 7 - % 28.0 1 No. - 2.0 - 1 6.3 - 1 2 5.9 4.0 - 4 - 25.0 1 6 5.9 11.1 - 21 - 1 1.9 14 ampico pearlymussel---- 440.0 111.1------1 5.9 220.0 3 5.611 3 220.0 5.9 1 111.1------440.0 1 pearlymussel------110.0------ampico fawnsfoot exas

______S ______T R T Y F W B G S S T P ______96 smooth pimpleback found the areas less acceptable, southern mapleleaf may have found conditions advanta- geous. Despite current environmental problems, southern mapleleaf remains one of the most abundant unionids statewide.

It should be noted that although current sampling efforts directed at living populations and more-recent shell deposits make an effort to evaluate all species and sizes present, prehistoric harvest may have focused on spe- cific species and sizes to the exclusion of others. Native Americans may have deliberately selected for smooth pimplebacks and against southern mapleleafs, for example. Abundance in archeological samples would then provide a misleading guide to relative abundance among mussel assemblages of the time. This issue typically remains unresolved unless harvested and associated natural shell deposits can be compared.

Environmentally, all of the unionid species present evolved in flowing waters (and all, except false spike, are known to tolerate impoundment). They are most often associated with gravel, stable sand, and mud substrates in permanent-water conditions. None are typical of ponds, oxbows, canals, temporary waters, or bottoms of silt, deep-shifting sands, or scoured cobble and bedrock. Most probably favor gravel riffles and bars.

Other unionid species known from the Central Brazos River Drainage (Table D-2; Howells et al. 1996) that were not present in the current set of samples may have been absent for a number of reasons. Harvest appears to have centered on stream and river areas and not ponds or temporary waters where species like pond mussel (Ligumia subrostrata) and pondhorns (Uniomerus spp.) would have occurred. Species from such habitats were not recovered. Unionid taxa taken were also thicker-shelled species that are more likely to survive than thinner- shelled taxa like giant floater (Pyganodon grandis) and fragile papershell (Leptodea fragilis) that often crack and disintegrate upon drying. These thinner shells may not have survived to the present as readily as heavier, thicker shells. Additionally, it is possible some species were deliberately selected against for a variety of rea- sons, including taste. Some, like washboard (Megalonaias nervosa), often occupy deeper waters and are less mobile. Both traits may then make them less likely to have been harvested than species found in shallower waters.

Although the valves and shells recovered from these sites were in relatively good condition compared to those from some other archeological sites, shell margins were worn in nearly every case. This, in turn, confounded attempts at assigning season of harvest to the specimens. Further, seasonality has yet to be demonstrated among unionid bivalves in Texas and growth-rest lines on shells of modern specimens suggest confident seasonality assignment locally is probably not possible (Texas Parks and Wildlife Department, unpublished data).

97 References Cited

Howells, R. G. 1995 Distributional surveys of freshwater bivalves in Texas: progress report for 1993. Texas Parks and Wildlife Department, Management Data Series 119, Austin.

Howells, R. G. 1996 Distributional surveys of freshwater bivalves in Texas: progress report for 1994. Texas Parks and Wildlife Department, Management Data Series 120, Austin.

Howells, R. G. 1997 Distributional surveys of freshwater bivalves in Texas: progress report for 1996. Texas Parks and Wildlife Department, Management Data Series 144, Austin.

Howells, R. G., C. M. Mather, and J. A. M. Bergmann 1997 Conservation Status of Selected Freshwater Mussels in Texas. In Conservation and Management of Freshwater Mussels II: Initiatives for the Future, edited by K. S. Cummings, A. C. Buchanan, C. A. Mayer, and T. J. Naimo, pp. 117–127. Proceedings of a UMRCC Symposium, 16–18 October 1995. Upper Mississippi Conservation Committee, Rock Island, Illinois.

Howells, R. G., R. W. Neck, and H. D. Murray 1996 Freshwater Mussels of Texas. Texas Parks and Wildlife Press, Austin.

Neck, R. W. 1982 A Review of Interactions Between Humans and Freshwater Mussels in Texas. In Proceedings of the Symposium on Recent Benthological Investigations in Texas and Adjacent States, edited by J. R. Davis, pp. 169–182. Texas Academy of Science, Austin.

Neves, R. J. 1993 A State of the Unionids Address. In Conservation and Management of Freshwater Mussels, edited by K. S. Cummings, A. C. Buchanan, and L. M. Koch, pp. 1–10. Proceeding of a UMRCC Symposium, 12–14 October 1992. Upper Mississippi River Conservation Committee, Rock Island, Illinois.

Strecker, J. 1931 The Distribution of Naiades or Pearly Fresh-Water Mussels of Texas. Baylor University Museum, Bulletin 2, Waco.

Williams, J. D., M. L. Warren, Jr., K. S. Cummings, J. L. Harris, and R. J. Neves 1993 Conservation status of freshwater mussels of the United States and Canada. Fisheries (Bethesda) 18(9):6–22.

98 Appendix E

Macrobotanical Analysis of Flotation Samples

J. Philip Dering Appendix E: Macrobotanical Analysis of Flotation Samples J. Philip Dering

Macrobotanical Analysis of Flotation Samples from the Little River Testing Project, 41MM340

This report presents results of the macrobotanical analysis of five flotation samples that were recovered during test excavations conducted at 41MM340. The botanical assemblage from the site is evaluated for its potential to yield significant botanical data.

Laboratory Methods Flotation samples are samples of archaeological sediment that have been floated in water to separate lighter charred plant remains from heavy material such as rock or caliche, and to remove lighter material such as clays/ silts that can be suspended in water and rinsed out of the sample. The light plant material is usually caught on fine-mesh material such as cheesecloth, or on very fine-mesh screen smaller than 0.450 mm. The sediment samples were floated by personnel from the Center for Archaeological Research at The University of Texas at San Antonio.

Standard archaeobotanical laboratory procedures were followed during analysis. The samples were first opened and dried in an herbarium dryer. Then each sample was sorted through a series of four nested geological screens with mesh sizes of 4-mm, 2-mm, 1-mm, and 0.450-mm. The material caught on all of the sieve levels, including the pan was scanned for floral parts, fruits, and seeds. Usually only the wood fragments caught on the 4-mm and 2-mm screens are separated and identified, because it is extremely difficult and time consuming to identify wood fragments smaller than 2 mm. Many times the diagnostic characteristics are not present in fragments smaller than 2–4-mm. The samples from 41MM340 were so small that it was decided to examine as much of the material caught on the 1-mm screen as possible. In this manner one more identification was secured.

Due to the poor preservation encountered at most open sites, only carbonized plant remains were considered for inclusion in the archaeological assemblage. Identification of all carbonized wood is accomplished by using the snap technique, examining them at 8 to 45 magnifications with a hand lens or a binocular dissecting microscope, and comparing them to samples in the herbarium at the Texas A&M University. Identifications are made using reference collections at Texas A&M University.

Results of the Macrobotanical Analysis Archaeobotanical Assemblage The highly reduced nature of the plant assemblage, both in quantity and in the size of the material, resulted in very few identifications. Three plant taxa were noted in the samples, oak wood, Ilex sp. (holly or haw) wood, and a possible pecan nut fragment (Table E-1). Plant parts noted in the samples included charred wood and nut fragments. All material labeled nut fragment refers to the outer pericarp of the nut.

The samples contained 17 wood fragments weighing approximately 0.3 grams, one cf. pecan fragment, and seven indeterminate nut fragments for a total of 25 charred plant fragments (Table E-2). Wood taxa represented in the samples were oak and Ilex sp. (possibly Ilex decidua or possum haw). Ilex sp. was noted in Lot 173, and

100 Table E-1. Plant taxa identified in the samples

Taxon Common Name Part cf. Carya illinoiensis Pecan Nut fragment Thin nutshell Nut fragment Quercus sp. Oak Wood Ilex sp. Holly, possum haw Wood oak was noted in Lots 75 and 222. Only two fragments of wood were larger than 3 mm and each of these larger fragments measured less than 4 mm in length. A single charred thin nut fragment, possibly pecan, was recovered from Lot 222. Other thin nut fragments too small to identify were noted in Lot 170 and Lot 173. All of the nut fragments were smaller than 3 mm in length. Thin nut fragments usually represent pecan, and thicker nut frag- ments a type of hickory. In this case all of the plant material was too small to provide positive identifications.

The few nut fragments recovered from Lots 170, 173, and 222 provide some evidence suggesting that nut processing may have occurred at the site. Although such a small quantity of fragments could easily have been incorporated into the fire incidentally with a load of fuel wood, nut processing at the site should be considered as a possibility. Nut processing is usually accomplished using groundstone tools including nutting stones/manos and metates. The ethnographic record indicates that hickory is usually processed into oil (Talalay et al. 1984). Processing techniques for pecans are poorly recorded, but pecan shell fragments have been recovered from sheltered sites in Val Verde County (Baker Cave).

According to Hall (2000) there is no direct archeological or ethnohistorical evidence for processing pecan into oil as is commonly recorded for hickory. Hickory nuts are usually pounded using manos and anvil stones and then boiled to separate the oil. The proximity to fire increases the opportunity for the charring of the thick hickory shell (pericarp). Pecans may be stored in the shell for at least a year in an adequate setting, and the Payaya Indians of southern Texas stored pecans in pits, possibly to compensate for the reduced pecan productiv- ity encountered after years of bounty (Hall 2000). Because pecans may not be processed around heating features like hickory, the chances for recovering charred pecan fragments may be reduced. Therefore the possible occur- rence of pecan shell at open archaeological sites is noteworthy.

Table E-2. Flotation samples, provenience, liters floated, and identifications

Sample Fraction and Lot # Block Unit Level Elevation Zone Plant Part Count Weight (g) Material Size 75 3 C 2 60-70 2 Oak Wood 4 <.1 LF 117 3 C 2 81/91 4 NICPR* Flecks < 1mm 173 3 C 1 94/104 6 Ilex sp. Wood 12 <.1 LF Indeterminate Thin nut frag. 1 <.1 LF (2 mm) 222 3 C 2 124/134 8 Pecan (cf.) Thin nut frag. 1 <.1 LF(3 mm) Oak Wood 1 <.1 LF (2 mm) Indeterminate Flecks LF Flecks < 1mm 170 1 B 6 50-60 Indeterminate Thin nut frag. 6 <.1 HF Fragments <2mm Total Fragments 25 * indicates samples with no identifiable carbonized plant remains. The first four samples are from 41MM340, the last sample is from 41MM341.

101 Conclusion Although the archaeobotanical assemblage from 41MM340 flotation samples is somewhat reduced, the botani- cal data have yielded some promising results. Nut fragments (pericarp fragments) were recovered from three of the five (60 percent) flotation samples examined. At least two nut types, thin (cf. pecan) and thick (cf. hickory), were identified in the assemblage.

The results from the wood charcoal study were not as promising. The wood assemblage was reduced and of the 25 wood fragments noted in the flotation samples only two were larger than 2 mm long. The rest of the charred wood material that I attempted to identify ranged between 1 and 2 mm in length, making the process of identifi- cation very difficult. One of the five samples contained only small flecks of charcoal of insufficient size for identification.

There are, however, a total of 76 (66 from 41MM340, 10 from 41MM341) charcoal samples that were collected by CAR-UTSA during the excavation that have not yet been submitted for analysis (Steve Tomka, personal communication, 2000). These samples contain larger quantities of both wood and nut fragments than were noted in the floatation samples. If subsequent studies are authorized, these samples should be considered for analysis because the floatation samples have demonstrated that the site has the potential to generate botanical data. Furthermore, other areas located within the site may contain pockets of improved plant preservation. As a result, it will be critical for archaeologists to monitor the recovery of plant remains from the site if further work is recommended, because pockets of improved preservation may be encountered.

References Cited

Hall, G. 2000 Pecan Food Potential in Prehistoric North America. Economic Botany 54(1):103–112.

Talalay, L., D. R. Keller, and P. J. Munson 1984 Hickory Nuts, Walnuts, Butternuts, and Hazelnuts: Observations and Experiments Relevant to Their Aboriginal Exploitation in Eastern North America. In Experiments and Observations on Aboriginal Wild Plant Food Utilization in Eastern North America, edited by P. J. Munson, pp. 338–359. Prehistory Research Series 6(2). Indiana Historical Society. Indianapolis, Indiana.

102 Appendix F

Pollen and Phytolith Results

John G. Jones Appendix F: Pollen and Phytolith Results John G. Jones

I have completed the pollen and phytolith evaluation from site 41MM340 and 41MM341. In all cases, pollen grains were noted, but were poorly preserved and would probably not provide valid information on past environ- mental conditions. Phytoliths likewise were poorly preserved reflecting the harsh environmental conditions in the site area. It is my opinion that pollen and phytolith analyses from these sites are not warranted.

Both fossil pollen and phytoliths suffer from degradation caused by oxidizing conditions. In the case of pollen, these conditions (i.e., high pH and cyclic wetting and drying) favor fungal and bacterial growth which can destroy pollen. Phytoliths are also negatively effected from oxidizing conditions, especially in areas with high pH values. Recognizing that the environmental conditions in Milam county well match the above criteria, con- servative processing techniques were employed in the extraction of the pollen and phytoliths.

Pollen was isolated using standard techniques employed at the Texas A&M University Palynology Laboratory. Exotic tracer spores were added to a standard volume of sediment (15 mls). The sample was then rinsed with HCl, HF, an acetolysis treatment, a wash in dilute KOH and a heavy density separation using ZnBr (Sp.G. 2.00). Residues were dehydrated in alcohol and were transferred to 1 dram vials for storage in glycerin. Slides were prepared using glycerin as a mounting media.

Phytoliths were also quantified (10 grams), rinsed in HCl, oxidized both Schulze’s solution (concentrated Nitric Acid and Potassium Chlorate) and then in 5% KOH. The lighter phytoliths were isolated from inorganic silicates using a heavy density separation with ZnBr (Sp.G. 2.38). The phytoliths were collected, rinsed, and mounted on permanent slides using Meltmount adhesive. Examinations of both pollen and phytoliths were made on a Jenaval compound stereomicroscope at 400x magnification.

Pollen was noted in all of the samples, and at least 13 different taxa were noted, including oaks, pines, junipers, ephedra, grasses, composites, thistles, goosefoot, lily, sedge, buttonbush and willow. However, nearly all grains were poorly preserved and most of the taxa recorded represent species that are especially abundant in the envi- ronment and are quite durable and easily recognized when degraded. Through the addition of exotic tracer spores, I was able to calculate the concentration values for the samples. These values ranged from 936 to 2250 fossil grains per ml of sediment. These values are considered quite low, and interpretations based on these pollen counts must be suspect. In short, although some fossil pollen remains in the archaeological sediments from these sites, little information could be obtained from a detailed pollen analysis of these sediments.

Phytoliths were extremely rare in the sediments from these sites. When noted, most phytoliths were highly degraded indicating that fossil biosilicates have been lost from the sediments. Conditions affecting phytolith preservation are imperfectly understood, but it is known that highly oxidizing conditions in areas with a high pH can result in total phytolith loss.

Although the examination of the microfossils was not encouraging, it is important to continue testing different regions as micro-edaphic conditions frequently allow for pollen and phytolith preservation in regions where one would not expect them to be found.

104 Appendix G

Gastropod Analysis

Richard W. Fullington Appendix G: Gastropod Analysis Richard W. Fullington, Ph.D.

Site 41MM340

Summary

A total of 1,640 gastropod shells were examined from six lots. Thirteen species were identified representing ten families (Table G-1). All shells were in excellent condition and coloration is still remarkably good. Literature records indicate that a number of the site species can be found today in the immediate area with the exception of Carychium mexicanum and Haplotrema concavum (Cheatum and Fullington, 1971, 1973; Fullington and Pratt 1974).

Climatically and environmentally, there appears to have been little change through the time period covered by the examined levels (50–134 cm bs) (Table G-2). The site area was heavily forested with a deep leaf litter floor — six of the site species prefer this type of habitat. There were open, grassy areas adjacent to these woods — six of the site species prefer this type of habitat. In summary, the area was a benched terrace with a riparian wood- land adjacent to a relatively slow moving stream. The one aquatic gastropod, Helisoma trivolvis lentum, prefers this type of aquatic habitat. Carychium mexicanum occurs in deep, constantly moist leaf litter and in moss alongside a stream.

The high numbers of Rabdotus dealbatus dealbatus found in Block 3, Zone 4, elevation 124–134 may indicate the presence of a trash pit. Terrestrial gastropods tend to be attracted by organic debris. The author has observed this phenomena many times during collection trips. Roadside dumps are a bonanza for snail collecting.

Methodology

Lots 64, 181 and 235 were sifted through ¾-inch screen and Lots 104, 173 and 222 were recovered from flota- tion processing and fine screening. The shells were examined through a standard field microscope (20x). All samples are from Block 3, Units C and B.

Data Analysis

The gastropod shells examined from Site 41MM340 appear to have been primarily deposited in situ and not from flood deposition. If this had been the case, the numbers of shells and species would have been much higher and there would have been more aquatic species found. There was little difference in the number of shells found between cultural and non-cultural zones. This indicates that the occupational sites were not heavily populated nor inhabited for lengthy periods of time (personal research).

The presence of the predatory snail, Haplotrema concavum, is indicative of a stable and long term gastropod community. It is a voracious feeder and occurs only where there are large and stable populations of snails.

106 Table G-1. Identification of gastropod shell remains Habitat Preferences recovered from 41MM340 Densely wooded forest with deep, moist leaf litter and Family Carychiidae fallen logs: Carychium mexicanum (Pilsbry) 6 Strobilops texasiana Family Pupillidae Helicodiscus parallelus Gastrocopta contracta (Say) 26 Hawaiia minuscula Gastrocopta pellucida hordeacella (Pilsbry) 2 Gastropcopta contracta Family Haplotrematidae Haplotrema concavum Haplotrema concavum (Say) 1

Family Zonitidae Open grassy areas to sparsely wooded areas: Hawaiia minuscula (Binney) 49 Rabdotus dealbatus dealbatus Zonitoides arboreus (Say) 13 Polygyra mooreana Family Helicinidae Praticolella berlandieriana Helicina orbiculata tropica (Say) 252 Helicina orbiculata tropica Gastrocopta pellucida hordeacella Family Endodontidae Helicodiscus parallelus (Say) 7 Zonitioides arboreus

Family Polygyridae Polygyra mooreana (Binney) 8 Moss adjacent to flowing water to very moist forest floor: Praticolella berlandieriana (Moricand) 2

Carychium mexicanum Family Bulimulidae Rabdotus dealbatus dealbatus (Say) 1,233 Potamon stream with emergent littoral emergent veg- Family Strobilopsidae etation: Strobilops texasiana (Pilsbry & Ferriss) 40 Helisom trivolvis lentum

Aquatic Gastropoda Family Planordidae (Aquatic) Helisoma trivolvis lentum (Say) 1 Total 1,640

107 Table G-2. Distribution of selected gastropod shell remains recovered from 41MM340

Gastropod Data by Lots Lot Block Unit Zone Level Elevation Species Number Taphonomy 64 3 C 2 6 50-60 Helicina orbiculata tropica 130 Subtotal 130

104 3 C 3 9 80-90 Helicina orbiculata tropica 6 104 3 C 3 9 80-90 Carychium mexicanum 3 104 3 C 3 9 80-90 Srobilops texasiana 21 104 3 C 3 9 80-90 Gastrocopta contracta 8 104 3 C 3 9 80-90 Gastrocopta pellucida hordeacella 2 104 3 C 3 9 80-90 Zonitoides arboeus 5 104 3 C 3 9 80-90 Hawaiia minuscula 18 104 3 C 3 9 80-90 Helicodiscus parallelus 3 Subtotal 66

173 3 C 4 10:11 94-104 Helicina orbiculata tropica 2 173 3 C 4 10:11 94-104 Carychium mexicanum 2 173 3 C 4 10:11 94-104 Strobilops texasiana 16 173 3 C 4 10:11 94-104 Gastrocopta contracta 14 173 3 C 4 10:11 94-104 Zonitoides arboreus 5 173 3 C 4 10:11 94-104 Hawaiia minuscula 20 173 3 C 4 10:11 94-104 Helicodiscus parallelus 4 Subtotal 63

181 3 B 4 10:11 94-104 Rabdotus dealbatus dealbatus 19 181 3 B 4 10:11 94-104 Helicina orbiculata tropica 68 181 3 B 4 10:11 94-104 Polygyra mooreana 7 181 3 B 4 10:11 94-104 Haplotrema concavum 1 Predatory land snail Subtotal 95

222 3 C 6 13:14 124-134 Rabdotus dealbatus dealbatus 4 222 3 C 6 13:14 124-134 Helicina orbiculata tropica 3 222 3 C 6 13:14 124-134 Carychium mexicanum 1 222 3 C 6 13:14 124-134 Strobilops texasiana 3 222 3 C 6 13:14 124-134 Gastrocopta contracta 4 222 3 C 6 13:14 124-134 Zonitoides arboreus 3 222 3 C 6 13:14 124-134 Hawaiia minuscula 11 Subtotal 29

235 3 B 4 13:14 124-134 Rabdotus dealbatus dealbatus 1,210 235 3 B 4 13:14 124-134 Helicina orbiculata tropica 43 235 3 B 4 13:14 124-134 Praticolella berlandieriana 2 235 3 B 4 13:14 124-134 Helisoma trivolvis lentum 1 Aquatic 235 3 B 4 13:14 124-134 Polygyra mooreana 1 Subtotal 1,257 Grand Total 1,640

108 Site 41MM341

Summary

A total of 195 gastropod shells were examined from 1 lot. Eleven species were identified representing eight families (Table G-3). All shells were in excellent condition and coloration is still remarkably good. Virtually all of the site species can be found today in the immediate area with the exception of Carychium mexicanum and Haplotrema concavum (Cheatum and Fullington, 1971, 1973; Fullington and Pratt 1974).

Climatically and environmentally, there appears to have been little change through the time period covered by the examined levels (50–60 cm bs) (Tables G-4). The Site area was heavily forested with a deep leaf litter floor — seven of the site species prefer this type of habitat. There were open, grassy areas adjacent to these woods — four of the site species prefer this type of habitat. In summary, the area was a benched terrace with a riparian woodland adjacent to a relative slow moving stream. Carychium mexicanum occurs in deep, constantly moist leaf litter and in moss alongside a stream. Exposed, wooded limestone outcroppings must have been nearby due to the presence of Anquispira alternata which inhabits this type of environment.

Methodology

Soil samples from Lot 34 were sifted through ¼-inch screen and shells from Lot 170 was recovered from flotation processing and fine screening. The shells were examined through a standard field microscope (20x). All samples are from Block 1, Units B and D.

Data Analysis

The gastropod shells examined from Site 41MM341 appear to have been primarily deposited in situ and not from flood deposition. If this had been the case, the numbers of shells and species would have been much higher and there would have been more aquatic species found. There was little difference in the number of shells found between cultural and non-cultural zones. This indicates that the occupational sites were not heavily populated nor inhabited for lengthy periods of time (personal research).

The presence of the predatory snail, Haplotrema concavum, is indicative of a stable and long term gastropod community. It is a voracious feeder and occurs only where there are large populations of snails.

Given the amount of data that was gleaned from a small sampling, site 41MM341 is well worth further investigation. The knowledge that the snail populations appear to have been deposited in situ provides an opportunity to further investigate the relationship of primitive people and their impact on the immediate environment. The Site also represents an opportunity to clarify some the site occupants activities through the terrestrial gastropods.

109 Table G-3. Identification of gastropod shell remains Habitat Preferences recovered from 41MM341 Densely wooded Forest with deep, moist leaf litter Family Carychiidae and fallen logs: Carychium mexicanum (Pilsbry) 63 Strobilops texasiana Family Pupilidae Helicodiscus parallelus Gastrocopta contracta (Say) 28 Hawaiia minuscula Gastrocopta pellucida hordeacella (Pilsbry) 7 Gastropcopta contracta Family Haplotrematidae Haplotrema concavum Haplotrema concavum (Say) 2 Euconulus fulvus

Family Zonitidae Hawaiia minuscula (Binney) 32 Open grassy areas to sparsely wooded areas: Zonitoides arboreus (Say) 11 Rabdotus dealbatus dealbatus Euconulus fulvus (Mueller) 1 Polygyra mooreana Family Helicinidae Praticolella berlandieriana Helicina orbiculata tropica (Say) 3 Helicina orbiculata tropica Gastrocopta pellucida hordeacella Family Endodontidae Anguispira alternata (Morse) 1 Zonitioides arboreus

Family Bulimulidae Moss adjacent to flowing water to very moist forest Rabdotus dealbatus dealbatus (Say) 5 floor:

Family Strobilopsidae Carychium mexicanum Strobilops texasiana (Pilsbry & Ferriss) 39 Total 195 Sloping hillsides with brush or trees (terrace slopes, limestone outcroppings):

Anguispira alternata

110 Table G-4. Distribution of selected gastropod shell remains recovered from 41MM341 Gastropod Data by Lots Lot Block Unit Level Elevation Species Number Taphonomy 34 1 D 6 50-60 Rabdotus dealbatus dealbatus 5 Larger & older appearing than 340 34 1 D 6 50-60 Helicina orbiculata tropica 3 34 1 D 6 50-60 Haplotrema concavum 2 Predatory land snail 34 1 D 6 50-60 Anguispira alternata 1 Orange still bright Total 11

170 1 B 6 50-60 Carychium mexicanum 63 170 1 B 6 50-60 Strobilops texasiana 39 170 1 B 6 50-60 Euconulus fulvus 1 170 1 B 6 50-60 Gastrocopta contracta 28 170 1 B 6 50-60 Gastrocopta pellucida hordeacella 7 170 1 B 6 50-60 Zonitoides arboreus 11 170 1 B 6 50-60 Hawaiia minuscula 32 Total 181

Future Site Investigation Recommendations

(Sites 41MM340, 41MM341)

Given the amount of data that was gleaned from a small sampling, Site 41MM340 is well worth further inves- tigation. The knowledge that the snail populations appear to have been deposited in situ provides an opportu- nity to further investigate the relationship of primitive people and their impact on the immediate environment. The site also represents an opportunity to clarify some the site occupant’s activities through the terrestrial gastropods.

Shell preservation is exceptionally good as is shell coloration and there appears to be a drying trend upwards through the site strata. This possibility also should be pursued with further site investigation.

Other future site investigations should include the following:

Survey of the existing gastropod fauna

Develop a single vertical column through the site strata to obtain a complete linear time line. Pollen samples should be taken from this column.

Develop a column through cultural layers and another column away from the cultural activity to test for human impact on the local fauna

The possible utilization of Rabdotus dealbatus as a food source by the site occupants

111 References Cited

Cheatum, E. P., and R. W. Fullington 1971 Bulletin 1 The Aquatic and Land Mollusca of Texas. Part One: The Recent and Pleistocene Members of the Gastropod Family Polygyridae in Texas. DMNH, Dallas. 1973 Bulletin 1 The Aquatic and Land Mollusca of Texas. Part Two: The Recent and Pleistocene Members of the Pupillidae and Urocoptidae (Gastropoda) in Texas. DMNH, Dallas.

Fullington, R. W., and W. L. Pratt, Jr., 1974 Bulletin 1 The Aquatic and Land Mollusca of Texas. Part Three: The Helicinidae, Carychiidae, Achatinidae, Bradybaenidae, Bulimulidae, Cionellidae, Haplotrematidae, Helicidae, Oreohelicidae, Spiraxidae, Streptaxidae, Strobilopsidae, Thysanophoridae, Valloniidae (Gastropoda) in Texas. DMNH, Dallas.

112 Appendix H

Diatom Paleoenvironmental Analysis of Sediments

Barbara Winsborough Appendix H: Diatom Paleoenvironmental Analysis of Sediments Barbara Winsborough

Introduction Diatoms are unicellular, eukaryotic, pigmented, photosynthetic algae distinguished by the possession of a cell wall that is heavily impregnated with silica. Some species are colonial, living in clusters, chains, or tubes, while others form zig-zag or star-shaped colonies. Many species produce extracellular mucilage pads, stalks, tubes, amorphous masses, strands, threads, or filaments that are composed of polysaccharide. Some diatom species are free-floating planktonic forms living much of their life suspended in the water column, and other species are benthic, either crawling unattached through sediments or spending their lives attached to a solid surface such as sand, pebble, rock, wood, rooted macrophyte, other algae, or floating mat. Most aerophilic species are motile and can glide over surfaces and through sediments for efficient light capture.

Diatoms can be found living in a wide variety of natural and man-made terrestrial and aquatic habitats, including seeps, wet walls, damp soil, caves, springs, streams, ponds, lakes, marshes, lagoons, estuaries, bogs, swamps, fens, ditches, canals, temporary pools, travertine accumulations, water and sewage treatment facilities, cooling towers, and hatcheries; on ice and snow; on turtles, whales, other mammals, invertebrates, and fish; and in estuaries, bays, oceans and seas. Most are cosmopolitan — found in many parts of the world under similar environmental conditions, and many species have predictable environmental requirements and pollution toler- ance. Therefore water quality directly effects diatom species composition. Diatoms can be readily identified to species, and a large body of information exists on the range of ecological tolerance of many common species. Because diatoms are sensitive indicators of water chemistry, habitat, and substrate; are often found in large numbers in sedimentary deposits; can be identified to species; and are cosmopolitan in distribution, they are well-suited for use in paleoenvironmental reconstruction.

Eight sediment samples and five fired clay samples (Table H-1) from two archaeological sites near the town of Cameron, in Milam County east-central Texas, were submitted to Winsborough Consulting for diatom paleoenvironmental analysis. Three of the eight sediment samples were modern samples collected to establish a local modern analog data set. Water bodies and water chemistry have probably changed little within the time frame under consideration in this investigation. These samples include a phytoplankton and bottom sludge sample from the Little River, and the top 1 cm of soil from a modern slough surface 60 m south of 41MM340 (Table H-1). The sediment samples were processed in the manner described below. The baked clay samples were first soaked in concentrated HCL to disaggregate them, then rinsed prior to being cleaned with the usual protocol described below.

Methods Observations are made on the diatom frustule after it is cleaned of its organic parts. A sample of sediment material measuring about a centimeter square was divided approximately in half, one half was cleaned and the other curated. Samples were cleaned of organic material and carbonate in preparation for microscopic analysis by boiling first in hydrogen peroxide and then in nitric acid. The oxidized, decalcified material was rinsed repeatedly until a pH of about 6–7 was reached. A few drops of the cleaned material was air-dried on to glass coverslips and mounted on glass slides using HYRAX ©, a resin with a high index of refraction. Slides were scanned at 1500x and 2000 diatoms were counted from each sample. Diatom relative abundance is determined from a count of 500 cells, whenever possible. The data are tabulated and the paleoecological significance of the individual species consid- ered in reconstruction of the environment.

114 Table H-1. Samples recovered from 41MM340 and 41MM341 examined for diatom content

Sample# Description 1 Modern phytoplankton from Little River at SH 77 crossing, collected 2-28-2000. 2 Bottom sludge sample from Little River at SH 77 crossing. collected 2-28-2000. 3 Top 1 cm of soil from modern slough located 60 m south of 41MM340, collected 2-28-2000. 4 Sediment from 41MM341, 2 cm below cultural zone. 5 Sediment from 41MM340, BHT A, south wall, zone 2. 6 Sediment from 41MM340, block 3, zone 4, north wall. 7 Sediment from 41MM340, block 3, zone 6, north wall. 8 Sediment from 41MM340, block 3, zone 8, north wall. 9 Fired clay nodules from 41MM340, lot 99, block 3, unit A, field zone 3-1, el. 70-80 cm. 10 Fired clay nodules from 41MM340, lot 116, block 3, unit C, zone 4-1, el. 84-94 cm. 11 Fired clay nodules from 41MM340, lot 185, block 3, unit A, zone 4-3, el. 94-104 cm. 12 Fired clay nodules from 41MM340, lot 229, block 3, unit A, zone 6-1, el. 124-134 cm. 13 Fired clay nodules from 41MM341, lot 67, block 2, unit B, zone 7, el. 60-70 cm.

Results Overall, there were 69 species of diatoms recorded from the samples. The largest number of taxa was in the modern benthic sludge sample from the Little River, which had 59 species in a count of 500 cells (Table H-2). This is a diverse population and probably represents an entire year or more of diatom accumulation, including various seasons and microhabitats. The diatoms encountered in the modern Little River sample are a mixture of benthic stream species and aerial taxa found in wetlands, temporary pools, on moss and wet rocks and in soils. They indicate a pH of circumneutral to slightly alkaline, shallow water, and a muddy substrate with some vegeta- tion. There were no clearly dominant species.

There were no diatoms in the phytoplankton sample. Phytoplankton populations often do not develop in small shallow streams with short residence times. Algal growth would be expected to be concentrated on the sediment surface and attached to the surface of vegetation on the bottom and sides of the stream.

The modern slough sample was diatomaceous, and had 10 species in it but the assemblage was heavily domi- nated by one diatom species, Craticula cuspidata. Craticula cuspidata is unusual among diatoms in that it lives in such habitats as temporary rainwater pools and saline pans in desert areas and forms internal ‘shells’ enabling it to remain alive in the dry state for several years. It may be an indicator of elevated salt concentrations.

The sediment samples (4, 5, 6, 7, and 8) were barren of diatoms, as well as phytoliths and pollen. If this material represents a flood deposit, it is not surprising that there would be very few diatoms if any in the material.

The fired clay nodules varied in their diatom content. Sample 9 was the most diatomaceous with 500 cells counted. It was also diverse, with 25 species recorded. These species are a mixture of aerophilic diatoms (found in soil and mud) and benthic aquatic species found attached to emergent vegetation and macroalgae. The assem- blage suggests that the paleoenvironment was a very shallow, muddy habitat with some vegetation. Most signifi- cantly, the diatoms in this sample were still attached in chains of up to 4 individuals, they were whole, and showed no signs of abrasion or breakage typical of an assemblage that has been transported and is just detrital. The large number of individuals found is also inconsistent with a detrital deposit. Samples 11 and 12 contained only a few aerophilic diatoms.

115 Table H-2. Identification of diatom samples recovered from Little River (modern) and sites 41MM340 and 41MM341

Sample # 2 3 9 11 12 Achnanthes clevei 2 Achnanthidium minutissium 14 Amphora montana 8 Amphora ovalis 331 Amphora pediculus 10 Amphora veneta 1 Bacillaria paradoxa 9 Caloneis bacillum 22 Cocconeis placentula 28 Craticula cuspidata 25 476 1 Cyclotella atomus 2 Cyclotella ocellata 1 Cymatopleura solea 2 Diadesmis confervacea 17 Diadesmis gallica 32 Diatoma vulgare 23 Diploneis parma 7 Encyonema silesiacum 71 Epithemia adnata 4 Epithemia turgida 4 Fragilaria ulna 18 1 Gomphonema angustum 6 Gomphonema clavatum 4 Gomphonema parvulum 13 19 Gyrosigma scalproides 24 Hantzschia amphioxys 2821 2 Hippodenta capitata 10 Hippodenta capitata var. hungarica 2 Luticola mutica 31302 Melosira varians 18 Meridion circulare 2 Navicula cryptocephala 311 Navicula cryptotenella 729 Navicula lenzii 1 Navicula menisculus 6 Navicula minima 66 Navicula pupula 4 Navicula rhynchocephala 86 Navicula salinarum 6 Navicula sanctaecrucis 21 Navicula subminuscula 9 Navicula symmetrica 12 Navicula texana 12 5 Navicula tripunctata 25 7 Navicula viridula 7 116 Table H-2. continued...

Sample # 2 3 9 11 12 Nitzschia acuminata 31 Nitzschia amphibia 24 3 58 Nitzschia angustata 6 Nitzschia compressa 2 Nitzschia constricta 5 Nitzschia dissipata 16 2 Nitzschia frustulum 14 Nitzschia inconspicua 15 4 Nitzschia levidensis 2 Nitzschia linearis 11 Nitzschia palea 10 111 Nitzschia solita Pinnularia cf. viridis 2 Plagiotropis lepidoptera var. proboscidea 11 Pleurosigma salinarum 1 Pseudostaurosira brevistriata 1 Reimeria sinuata 15 Rhoicosphena curvata 2 Rhopalodia brebissonii 452 Rhopalodia gibba 35 Staurosira construens 17 Surirella brebissonii 7 Surirella minuta 5 Surirella ovalis 5 Total 500 500 500 6 2 Note: Sample #s 1, 4, 5, 6, 7, 8, 10, and 13 were barren of diatoms.

Conclusions The most notable conclusion is that Sample 9, which represents fired clay nodules associated with archaeologi- cal hearth features, contains a concentration of diatoms that may have been transported to the site from a nearby stream during food preparation, possibly attached to mussels or other edible material. This provides confirma- tion that the feature was indeed a hearth impacted by human activities.

117

Appendix I

Amino Acid Racemization Values for Snail and Mussel Shell Appendix I : Amino Acid Racemization Values for Snail and Mussel Shell Data supplied by Texas Department of Transportation

Table I-1. Snail Shell Ratios and ASP/GLU Ratios Lab # Lot # Zone Polygyra mooreana Praticolella berlandieriana Aspartic acid/Glutamic acid CH-195 78 2 0.055 1.081 CH-196 78 2 0.0385 1.105 CH-197 78 2 0.0432 1.068 CH-198 78 2 0.0415 1.153 CH-199 78 2 0.0331 1.088 CH-200 78 2 0.0294 1.100 CH-201 78 2 0.0643 1.116 CH-202 78 2 0.0711 1.097 CH-163 150 4 0.0393 1.179 CH-164 150 4 0.065 1.073 CH-165 150 4 0.0593 1.061 CH-166 150 4 0.0403 1.056 CH-167 150 4 0.0507 1.147 CH-168 150 4 0.0359 1.108 CH-169 150 4 0.0663 1.006 CH-170 150 4 0.0449 1.110 CH-187 185 6 0.0587 1.150 CH-188 185 6 0.0405 1.114 CH-189 185 6 0.0286 1.090 CH-190 185 6 0.0379 1.113 CH-191 185 6 0.0603 1.116 CH-192 185 6 0.0596 1.067 CH-193 185 6 0.0358 1.201 CH-194 185 6 0.0185 1.264 CH-155 229 8 0.0598 1.060 CH-156 229 8 0.0279 1.185 CH-157 229 8 0.0536 0.943 CH-158 229 8 0.0291 1.115 CH-159 229 8 0.0716 1.042 CH-160 229 8 0.0401 1.022 CH-161 229 8 0.0555 1.001 CH-162 229 8 0.0611 1.024

120 Table I-2. Mussel Shell A/I Ratios and ASP/GLU Ratios

Lab # Lot # Zone A/I Ratio Aspartic acid/Glutamic acid Species CH-218 78 2 0.0605 1.900 Tritogonia verrucosa CH-215 78 2 0.0708 1.870 Amblema plicata CH-216 78 2 0.0811 1.830 Amblema plicata CH-217 78 2 0.0920 1.820 Tritogonia verrucosa CH-211 78 2 0.0957 1.880 Amblema plicata CH-212 78 2 0.1040 1.750 Amblema plicata CH-213 78 2 0.1070 1.900 Amblema plicata CH-214 78 2 0.3550 1.120 Amblema plicata CH-183 150 4 0.0487 1.180 Amblema plicata CH-182 150 4 0.0487 3.030 Amblema plicata CH-180 150 4 0.0655 1.880 Amblema plicata CH-185 150 4 0.0666 1.990 Amblema plicata CH-179 150 4 0.0759 2.010 Amblema plicata CH-181 150 4 0.0800 1.830 Amblema plicata CH-184 150 4 0.2240 1.320 Amblema plicata CH-186 150 4 0.2310 1.420 Amblema plicata CH-207 185 6 0.0569 2.010 Amblema plicata CH-208 185 6 0.0762 2.160 Amblema plicata CH-203 185 6 0.0785 2.000 Amblema plicata CH-204 185 6 0.0819 2.020 Amblema plicata CH-206 185 6 0.0886 2.190 Amblema plicata CH-205 185 6 0.0959 1.890 Amblema plicata CH-210 185 6 0.0965 2.010 Tritogonia verrucosa CH-209 185 6 0.9850 0.310 Tritogonia verrucosa CH-175 229 8 0.0557 1.250 Amblema plicata CH-171 229 8 0.0651 2.080 Amblema plicata CH-173 229 8 0.0812 1.150 Amblema plicata CH-176 229 8 0.0943 1.000 Amblema plicata CH-172 229 8 0.1720 1.750 Amblema plicata CH-177 229 8 0.2240 1.340 Amblema plicata CH-174 229 8 0.2960 0.640 Amblema plicata CH-178 229 8 1.0100 0.160 Tritogonia verrucosa

121

Appendix J

Prehistoric Ceramic Sherds from 41MM341

Timothy K. Perttula Appendix J: Prehistoric Ceramic Sherds from 41MM341 Timothy K. Perttula

Analysis of the Ceramic Sherds Thirteen prehistoric ceramic sherds have been recovered from 41MM341 on the Little River in Milam County, Texas, during recent archeological investigations by the Center for Archaeological Research at The University of Texas at San Antonio. The prehistoric component dates about 1050 BP. The sherds are from 0–30 cm bs in Block 1, Unit A (n=2), Unit B (n=9), and BHT 24 (n=2).

The sherds are from two different vessels. The first vessel (n=11 sherds, including five sherdlets from Unit B) is represented by all the sherds from Unit A and BHT 24, and seven of the sherds in Unit B; the sherdlets are less than 1 cm in length and width. This vessel has finely crushed and burned bone temper inclusions, and it has been fired in a reducing or low oxygen environment (cf. Teltser 1993:Figure 2B). The sherd cores are black. The sherds are from the body of a thick-walled vessel, probably a jar; the body wall thickness ranges from 6.9–9.4 mm, with a mean of 8.8 ± 0.6 mm. Two of the body sherds have a smoothed exterior surface, and one (from BHT 24) has remnants of an organic residue from use of the vessel for cooking foodstuffs

The second vessel includes two sherds from Unit B. The sherds are from a carinated bowl with bone and grog temper, and the clay paste is sandy. The carinated bowl had been fired in a reducing environment, but cooled in a high oxygen environment (cf. Teltser 1993:Figure 2F). Body wall thickness ranges from 7.9–8.4 mm, with a mean of 8.15 ± 0.25 mm. One of the carinated bowl sherds has a single broad horizontal incised line above the carination, and it is likely the decorative element consists of multiple horizontal lines on the rim of the vessel.

Discussion Although the ceramic sherds from 41MM341 are tempered with bone, the limited amount of bone in the paste (less than 10 percent of the paste matrix), the thick body walls (i.e., almost twice as thick as bone-tempered Leon Plain from Central Texas sites), and the absence of well-burnished exterior and/or interior surfaces indicate that the bone-tempered wares from the site are not from post-A.D. 1250 Leon Plain vessels. Whether it is a locally- made vessel is unknown at the present time, although this seems a reasonable conclusion. The occurrence of grog temper in the incised carinated bowl from 41MM341 strongly suggests that this vessel was made in the Caddoan area of northeastern Texas. Grog temper is widely used in Caddoan vessels of all time periods, and the carinated bowl form also appears to be a Caddoan innovation in Texas. To ascertain the manufacturing locale of the two different vessels, I recommend that a sherd from each vessel be analyzed by instrumental neutron activation analysis (INAA) and petrographic analysis. The current INAA and petrographic databases for Caddoan and Central Texas ceramics are sufficiently robust (see Perttula 1999) to be able to determine the likely compo- sitional origin of the two vessels.

Similar bone and/or grog-tempered sherds have been reported from Post Oak Savanna sites in east central Texas, particularly in the Navasota River basin, although prehistoric ceramics are not abundant in contexts dating from 1250–250 BP (Fields 1995:314–318, 323–324). At sites in Grimes County, for example, plain and decorated grog and bone-tempered sherds, and plain sherds with bone temper and/or sandy pastes; the decorated sherds have stylistic affiliations with Caddoan wares as well as San Jacinto wares made on the upper Texas coast (Rogers 1995). Turpin and Carpenter (1994:17) recovered thick (9.0 mm) and undecorated grog-tempered sandy paste

124 sherds at 41RT285 on Walnut Creek, near its confluence with the Brazos River, but in a context postdating A.D. 1250. At site 41RT510 in the Navasota River basin (Perttula n.d.), grog, grog-sandy paste, and bone-tempered sherds from plain jars and bowls are relatively abundant (n=36), and the vessels are relatively thick (6.6–10.5 mm). Rims are direct with rounded or flattened lips, and the sherds are from vessels primarily reduced during firing (83%). The few decorated sherds in the 41RT510 collection –horizontal and diagonal incised lines and fingernail punctated elements– suggest that the ceramics from this site date from ca. A.D. 900–1200, roughly contemporaneous with the ceramics from 41MM341.

References Cited

Fields, R. C. 1995 The Archeology of the Post Oak Savannah of East Central Texas. Bulletin of the Texas Archeological Society 66:301–330.

Perttula, T. K. 1999 Ceramic Evidence for Prehistoric Caddoan Trade and Exchange. Mini-NSF Grant Proposal on file, Research Reactor Center, University of Missouri, Columbia. n.d. The Ceramic and Lithic Assemblage from a Late Prehistoric Site (41RT510) in the Post Oak Savanna, Robertson County, Texas. MS in preparation.

Rogers, R. 1995 Archeological Excavations at Prehistoric Sites 41GM166, 41GM281, and 41GM282 at the Gibbons Creek Lignite Mine, Grimes County, Texas. Document No. 940611. Espey, Huston & Associates, Inc., Austin.

Teltser, P. A. 1993 An Analytic Strategy for Studying Assemblage-Scale Ceramic Variation: A Case Study from South- east Missouri. American Antiquity 58(3):530–543.

Turpin, J., and S. M. Carpenter 1994 Continuing Cultural Resource Management at Calvert Mine: 1993 Prehistoric Site Testing, Robertson County. Technical Series No. 42. Texas Archeological Research Laboratory, The University of Texas at Austin.

125 The following information is provided in accordance with the General Rules of Practice and Procedure, Chapter 41.11 (Investigative Reports), Texas Antiquities Committee:

1. Type of investigation: Phase II shovel testing

2. Project name: Little River Testing

3. County: Milam

4. Principal investigator(s): Robert J. Hard and Raymond P. Mauldin

5. Name and location of sponsoring agency: Texas Department of Transportation, Environmental Affairs Division, 125 East 11th Street, Austin, TX 78701-2843.

6. Texas Antiquities Permit No.: 2319

7. Published by the Center for Archaeological Research, The University of Texas at San Antonio, 6900 N. Loop 1604 W., San Antonio, Texas 78249-0658, 2001

A list of publications offered by the Center for Archaeological Research is available. Call (210) 458-4378; write to the Center for Archaeological Research, The University of Texas at San Antonio, 6900 N. Loop 1604 W., San Antonio, Texas 78249-0658; e-mail to [email protected]; or visit CAR’s web site at http://car.utsa.edu.